Critical ivector bugfix and version 0.3.15

Thanks to Markus Elfring

.gitignore

@@ -5,6 +5,9 @@
 # Java class files
 *.class
 
+# Object files
+*.o
+
 # Gradle files
 .gradle/
 build/
@@ -24,14 +27,37 @@ wheelhouse
 __pycache__
 *.egg-info
 python/dist
+python/vosk/*.cc
+python/vosk/*.c
+python/vosk/*.h
+python/vosk/*.i
+python/vosk/vosk.py
+python/vosk/vosk_wrap.cpp
+python/test/db
+python/test/hyp
+python/test/model
+python/test/ref
+python/test/result.txt
+python/test/wav.scp
 
 # Java
 *.so
 java/org
-java/model-en
 java/*.cc
+java/model-spk/
+java/model/
 
 # CSharp
 csharp/gen
 csharp/*.exe
-csharp/*.cc
+csharp/*.c
+csharp/model/
+csharp/test.wav
+
+# Node
+nodejs/vosk_wrap.cc
+nodejs/example/model
+nodejs/example/test.wav
+nodejs/node_modules
+nodejs/package-lock.json
+nodejs/build

README.md

@@ -1,114 +1,25 @@
-[![Build Status](https://travis-ci.com/alphacep/vosk-api.svg?branch=master)](https://travis-ci.com/alphacep/vosk-api)
+# About
 
-Language bindings for Vosk and Kaldi to access speech recognition from various languages and on various platforms
+Vosk is an offline open source speech recognition toolkit. It enables
+speech recognition models for 16 languages and dialects - English, Indian
+English, German, French, Spanish, Portuguese, Chinese, Russian, Turkish,
+Vietnamese, Italian, Dutch, Catalan, Arabic, Greek, Farsi.
 
-  * Python on Linux, Windows and RPi
-  * Node
-  * Android
-  * iOS
+Vosk models are small (50 Mb) but provide continuous large vocabulary
+transcription, zero-latency response with streaming API, reconfigurable
+vocabulary and speaker identification.
 
-## Android build
+Speech recognition bindings implemented for various programming languages
+like Python, Java, Node.JS, C#, C++ and others.
 
-```
-cd android
-gradle build
-```
+Vosk supplies speech recognition for chatbots, smart home appliances,
+virtual assistants. It can also create subtitles for movies,
+transcription for lectures and interviews.
 
-Please note that medium blog post about 64-bit is not relevant anymore, the script builds x86, arm64 and armv7 libraries automatically without any modifications.
+Vosk scales from small devices like Raspberry Pi or Android smartphone to
+big clusters.
 
-## Python installation from Pypi
+# Documentation
 
-The easiest way to install vosk api is with pip. You do not have to compile anything. We currently support only Linux on x86_64 and Raspberry Pi. Other systems (windows, mac) will come soon.
-
-Make sure you have newer pip and python:
-
-  * Python version >= 3.4
-  * pip version >= 19.0
-
-Uprade python and pip if needed. Then install vosk on Linux with a simple command
-
-```
-pip3 install vosk
-```
-
-## Compilation from source
-
-If you still want to build from scratch, you can compile Kaldi and Vosk yourself. The compilation is straightforward but might be a little confusing for newbie. In case you want to follow this, please watch the errors.
-
-#### Kaldi compilation for local python, node and java modules
-
-```
-git clone -b lookahead --single-branch https://github.com/alphacep/kaldi
-cd kaldi/tools
-make
-```
-
-install all dependencies and repeat `make` if needed
-
-```
-extras/install_openblas.sh
-cd ../src
-./configure --mathlib=OPENBLAS --shared --use-cuda=no
-make -j 10
-```
-
-#### Python module build
-
-Then build the python module
-
-```
-export KALDI_ROOT=<KALDI_ROOT>
-cd python
-python3 setup.py install
-```
-
-#### Java example API build
-
-Or Java
-
-```
-cd java && KALDI_ROOT=<KALDI_ROOT> make
-wget https://github.com/alphacep/kaldi-android-demo/releases/download/2020-01/alphacep-model-android-en-us-0.3.tar.gz
-tar xf alphacep-model-android-en-us-0.3.tar.gz 
-mv alphacep-model-android-en-us-0.3 model
-make run
-```
-
-#### C# build
-
-Or C#
-
-```
-cd csharp && KALDI_ROOT=<KALDI_ROOT> make
-wget https://github.com/alphacep/kaldi-android-demo/releases/download/2020-01/alphacep-model-android-en-us-0.3.tar.gz
-tar xf alphacep-model-android-en-us-0.3.tar.gz 
-mv alphacep-model-android-en-us-0.3 model
-mono test.exe
-```
-
-## Running the example code with python
-
-Run like this:
-
-```
-cd vosk-api/python/example
-wget https://github.com/alphacep/kaldi-android-demo/releases/download/2020-01/alphacep-model-android-en-us-0.3.tar.gz
-tar xf alphacep-model-android-en-us-0.3.tar.gz 
-mv alphacep-model-android-en-us-0.3 model
-python3 ./test_simple.py test.wav
-```
-
-There are models for other languages (English, Chinese, Spanish, Portuguese, German, French, Russian) available too at https://github.com/alphacep/kaldi-android-demo/releases
-
-To run with your audio file make sure it has proper format - PCM 16khz 16bit mono, otherwise decoding will not work.
-
-You can find other examples of using a microphone, decoding with a fixed small vocabulary or speaker identification setup in  [python/example subfolder](https://github.com/alphacep/vosk-api/tree/master/python/example)
-
-## Contact Us
-
-If you have any questions, feel free to
-
-   * Post an issue here on github
-   * Send us an e-mail at [contact@alphacephei.com](mailto:contact@alphacephei.com)
-   * Join our group dedicated to speech recognition on Telegram [@speech_recognition](https://t.me/speech_recognition)
-   
+For installation instructions, examples and documentation visit [Vosk
+Website](https://alphacephei.com/vosk).

android/CMakeLists.txt

@@ -8,6 +8,9 @@ set(KALDI_SUFFIX "arm_32")
 elseif ("x${ANDROID_ABI}" STREQUAL "xarm64-v8a")
 set(OPENBLAS_ARCH "armv8")
 set(KALDI_SUFFIX "arm_64")
+elseif ("x${ANDROID_ABI}" STREQUAL "xx86")
+set(OPENBLAS_ARCH "atom")
+set(KALDI_SUFFIX "x86")
 else ("x${ANDROID_ABI}" STREQUAL "xarmeabi-v7a")
 set(OPENBLAS_ARCH "atom")
 set(KALDI_SUFFIX "x86_64")
@@ -19,22 +22,26 @@ set(LIB_ROOT "${PROJECT_SOURCE_DIR}/build/kaldi_${KALDI_SUFFIX}/local")
 set(API_SOURCES
 "${PROJECT_SOURCE_DIR}/../src/kaldi_recognizer.cc"
 "${PROJECT_SOURCE_DIR}/../src/kaldi_recognizer.h"
+"${PROJECT_SOURCE_DIR}/../src/language_model.cc"
+"${PROJECT_SOURCE_DIR}/../src/language_model.h"
 "${PROJECT_SOURCE_DIR}/../src/model.cc"
 "${PROJECT_SOURCE_DIR}/../src/model.h"
 "${PROJECT_SOURCE_DIR}/../src/spk_model.cc"
 "${PROJECT_SOURCE_DIR}/../src/spk_model.h"
+"${PROJECT_SOURCE_DIR}/../src/vosk_api.cc"
+"${PROJECT_SOURCE_DIR}/../src/vosk_api.h"
 )
 
 set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -O3 -DFST_NO_DYNAMIC_LINKING")
 
-add_library( kaldi_jni SHARED
+add_library( vosk_jni SHARED
              build/generated-src/cpp/vosk_wrap.cc
              ${API_SOURCES}
 )
 
 include_directories("${PROJECT_SOURCE_DIR}/../src" "build/kaldi_${KALDI_SUFFIX}/kaldi/src" "build/kaldi_${KALDI_SUFFIX}/local/include")
 
-target_link_libraries( kaldi_jni
+target_link_libraries( vosk_jni
              ${KALDI_ROOT}/src/online2/kaldi-online2.a
              ${KALDI_ROOT}/src/decoder/kaldi-decoder.a
              ${KALDI_ROOT}/src/ivector/kaldi-ivector.a
@@ -43,6 +50,7 @@ target_link_libraries( kaldi_jni
              ${KALDI_ROOT}/src/tree/kaldi-tree.a
              ${KALDI_ROOT}/src/feat/kaldi-feat.a
              ${KALDI_ROOT}/src/lat/kaldi-lat.a
+             ${KALDI_ROOT}/src/lm/kaldi-lm.a
              ${KALDI_ROOT}/src/hmm/kaldi-hmm.a
              ${KALDI_ROOT}/src/transform/kaldi-transform.a
              ${KALDI_ROOT}/src/cudamatrix/kaldi-cudamatrix.a

android/build-kaldi.sh

@@ -31,36 +31,38 @@ fi
 
 set -x
 
+OS_NAME=`echo $(uname -s) | tr '[:upper:]' '[:lower:]'`
 ANDROID_NDK_HOME=$ANDROID_SDK_HOME/ndk-bundle
-ANDROID_TOOLCHAIN_PATH=$ANDROID_NDK_HOME/toolchains/llvm/prebuilt/linux-x86_64
+ANDROID_TOOLCHAIN_PATH=$ANDROID_NDK_HOME/toolchains/llvm/prebuilt/${OS_NAME}-x86_64
+WORKDIR_X86=`pwd`/build/kaldi_x86
 WORKDIR_X86_64=`pwd`/build/kaldi_x86_64
 WORKDIR_ARM32=`pwd`/build/kaldi_arm_32
 WORKDIR_ARM64=`pwd`/build/kaldi_arm_64
-PATH=$PATH:$ANDROID_NDK_HOME/toolchains/llvm/prebuilt/linux-x86_64/bin
+PATH=$PATH:$ANDROID_NDK_HOME/toolchains/llvm/prebuilt/${OS_NAME}-x86_64/bin
 OPENFST_VERSION=1.6.7
 
-mkdir -p $WORKDIR_ARM64/local/lib $WORKDIR_ARM32/local/lib $WORKDIR_X86_64/local/lib
+mkdir -p $WORKDIR_ARM64/local/lib $WORKDIR_ARM32/local/lib $WORKDIR_X86_64/local/lib $WORKDIR_X86/local/lib
 
 # Build standalone CLAPACK since gfortran is missing
 cd build
 git clone https://github.com/simonlynen/android_libs
 cd android_libs/lapack
-sed -i 's/APP_STL := gnustl_static/APP_STL := c++_static/g' jni/Application.mk && \
-sed -i 's/android-10/android-21/g' project.properties && \
-sed -i 's/APP_ABI := armeabi armeabi-v7a/APP_ABI := armeabi-v7a arm64-v8a x86_64/g' jni/Application.mk && \
-sed -i 's/LOCAL_MODULE:= testlapack/#LOCAL_MODULE:= testlapack/g' jni/Android.mk && \
-sed -i 's/LOCAL_SRC_FILES:= testclapack.cpp/#LOCAL_SRC_FILES:= testclapack.cpp/g' jni/Android.mk && \
-sed -i 's/LOCAL_STATIC_LIBRARIES := lapack/#LOCAL_STATIC_LIBRARIES := lapack/g' jni/Android.mk && \
-sed -i 's/include $(BUILD_SHARED_LIBRARY)/#include $(BUILD_SHARED_LIBRARY)/g' jni/Android.mk && \
+sed -i.bak -e 's/APP_STL := gnustl_static/APP_STL := c++_static/g' jni/Application.mk && \
+sed -i.bak -e 's/android-10/android-21/g' project.properties && \
+sed -i.bak -e 's/APP_ABI := armeabi armeabi-v7a/APP_ABI := armeabi-v7a arm64-v8a x86_64 x86/g' jni/Application.mk && \
+sed -i.bak -e 's/LOCAL_MODULE:= testlapack/#LOCAL_MODULE:= testlapack/g' jni/Android.mk && \
+sed -i.bak -e 's/LOCAL_SRC_FILES:= testclapack.cpp/#LOCAL_SRC_FILES:= testclapack.cpp/g' jni/Android.mk && \
+sed -i.bak -e 's/LOCAL_STATIC_LIBRARIES := lapack/#LOCAL_STATIC_LIBRARIES := lapack/g' jni/Android.mk && \
+sed -i.bak -e 's/include $(BUILD_SHARED_LIBRARY)/#include $(BUILD_SHARED_LIBRARY)/g' jni/Android.mk && \
 ${ANDROID_NDK_HOME}/ndk-build && \
 cp obj/local/armeabi-v7a/*.a ${WORKDIR_ARM32}/local/lib && \
 cp obj/local/arm64-v8a/*.a ${WORKDIR_ARM64}/local/lib
 cp obj/local/x86_64/*.a ${WORKDIR_X86_64}/local/lib
+cp obj/local/x86/*.a ${WORKDIR_X86}/local/lib
 
 # Architecture-specific part
 
-
-for arch in arm32 arm64 x86_64; do
+for arch in arm32 arm64 x86_64 x86; do
 #for arch in x86_64; do
 
 case $arch in
@@ -91,6 +93,15 @@ case $arch in
           CXX=x86_64-linux-android21-clang++
           ARCHFLAGS=""
           ;;
+    x86)
+          BLAS_ARCH=ATOM
+          WORKDIR=$WORKDIR_X86
+          HOST=i686-linux-android
+          AR=i686-linux-android-ar
+          CC=i686-linux-android21-clang
+          CXX=i686-linux-android21-clang++
+          ARCHFLAGS=""
+          ;;
 esac
 
 # openblas first
@@ -101,12 +112,9 @@ make -C OpenBLAS install PREFIX=$WORKDIR/local
 
 # tools directory --> we'll only compile OpenFST
 cd $WORKDIR
-wget -c -T 10 -t 1 http://www.openfst.org/twiki/pub/FST/FstDownload/openfst-${OPENFST_VERSION}.tar.gz || \
-wget -c -T 10 -t 3 http://www.openslr.org/resources/2/openfst-${OPENFST_VERSION}.tar.gz
-
-tar -zxvf openfst-${OPENFST_VERSION}.tar.gz
-cd openfst-${OPENFST_VERSION}
-
+git clone https://github.com/alphacep/openfst
+cd openfst
+autoreconf -i
 CXX=$CXX CXXFLAGS="$ARCHFLAGS -O3 -DFST_NO_DYNAMIC_LINKING" ./configure --prefix=${WORKDIR}/local \
     --enable-shared --enable-static --with-pic --disable-bin \
     --enable-lookahead-fsts --enable-ngram-fsts --host=$HOST --build=x86-linux-gnu
@@ -117,6 +125,9 @@ make install
 cd $WORKDIR
 git clone -b android-mix --single-branch https://github.com/alphacep/kaldi
 cd $WORKDIR/kaldi/src
+if [ "`uname`" == "Darwin"  ]; then
+  sed -i.bak -e 's/libfst.dylib/libfst.a/' configure
+fi
 
 CXX=$CXX CXXFLAGS="$ARCHFLAGS -O3 -DFST_NO_DYNAMIC_LINKING" ./configure --use-cuda=no \
     --mathlib=OPENBLAS --shared \
@@ -125,6 +136,6 @@ CXX=$CXX CXXFLAGS="$ARCHFLAGS -O3 -DFST_NO_DYNAMIC_LINKING" ./configure --use-cu
     --fst-root=${WORKDIR}/local --fst-version=${OPENFST_VERSION}
 
 make -j 8 depend
-make -j 8 online2
+make -j 8 online2 lm
 
 done

android/build.gradle

@@ -5,9 +5,17 @@ buildscript {
     }
     dependencies {
         classpath 'com.android.tools.build:gradle:3.5.3'
+        classpath 'com.jfrog.bintray.gradle:gradle-bintray-plugin:1.8.5'
     }
 }
 
+plugins {
+    id "com.jfrog.bintray" version "1.8.5"
+}
+
+def archiveName = "vosk-android"
+def libVersion = "0.3.15"
+
 allprojects {
   repositories {
       google()
@@ -16,22 +24,23 @@ allprojects {
 }
 
 apply plugin: 'com.android.library'
+apply plugin: 'maven-publish'
 
 android {
     compileSdkVersion 29
     defaultConfig {
         minSdkVersion 21
         targetSdkVersion 29
-        versionCode 5
-        versionName "5.2"
-        setProperty("archivesBaseName", "kaldi-android-$versionName")
+        versionCode 6
+        versionName = libVersion
+        archivesBaseName = archiveName
         externalNativeBuild {
             cmake {
                 arguments "-DCMAKE_VERBOSE_MAKEFILE=ON", "-DANDROID_ARM_NEON=TRUE", "-DCMAKE_CXX_FLAGS_RELEASE=-O3"
             }
         }
         ndk {
-            abiFilters 'armeabi-v7a', 'arm64-v8a', 'x86_64'
+            abiFilters 'armeabi-v7a', 'arm64-v8a', 'x86_64', 'x86'
         }
     }
     sourceSets {
@@ -46,6 +55,39 @@ android {
     }
 }
 
+Properties properties = new Properties()
+properties.load(project.rootProject.file('local.properties').newDataInputStream())
+
+bintray {
+    user = properties.getProperty("bintray.user")
+    key = properties.getProperty("bintray.apikey")
+    pkg {
+        repo = 'vosk'
+        name = 'vosk-android'
+        userOrg = "alphacep"
+        licenses = ['Apache2.0']
+        websiteUrl = 'https://github.com/alphacep/vosk-api'
+        issueTrackerUrl = 'https://github.com/alphacep/vosk-api/issues'
+        vcsUrl = 'https://github.com/alphacep/vosk-api'
+        version {
+            name = libVersion
+            vcsTag = libVersion
+        }
+    }
+    publications = ['aar']
+}
+
+publishing {
+    publications {
+        aar(MavenPublication) {
+            groupId 'com.alphacep'
+            artifactId archiveName
+            version libVersion
+            artifact("$buildDir/outputs/aar/$archiveName-release.aar")
+        }
+    }
+}
+
 task swig {
     doLast {
         mkdir 'build/generated-src/java'

android/src/main/AndroidManifest.xml

@@ -1,3 +1,3 @@
 <?xml version="1.0" encoding="utf-8"?>
-<manifest xmlns:android="http://schemas.android.com/apk/res/android" package="edu.cmu.pocketsphinx">
+<manifest xmlns:android="http://schemas.android.com/apk/res/android" package="org.kaldi">
 </manifest>

android/src/main/java/org/kaldi/Assets.java

@@ -241,10 +241,6 @@ public class Assets {
             if (!items.get(path).equals(externalItems.get(path))
                     || !(new File(externalDir, path).exists()))
                 newItems.add(path);
-            else
-                Log.i(TAG,
-                        String.format("Skipping asset %s: checksums are equal", path));
-
         }
 
         unusedItems.addAll(externalItems.keySet());
@@ -252,13 +248,11 @@ public class Assets {
 
         for (String path : newItems) {
             File file = copy(path);
-            Log.i(TAG, String.format("Copying asset %s to %s", path, file));
         }
 
         for (String path : unusedItems) {
             File file = new File(externalDir, path);
             file.delete();
-            Log.i(TAG, String.format("Removing asset %s", file));
         }
 
         updateItemList(items);

android/src/main/java/org/kaldi/SpeechService.java

@@ -29,19 +29,18 @@ import android.os.Looper;
 import android.util.Log;
 
 /**
- * Main class to access recognizer functions. After configuration this class
- * starts a listener thread which records the data and recognizes it using
- * Pocketsphinx engine. Recognition events are passed to a client using
+ * Service that records audio in a thread, passes it to a recognizer and emits 
+ * recognition results. Recognition events are passed to a client using
  * {@link RecognitionListener}
- * 
+ *
  */
-public class SpeechRecognizer {
+public class SpeechService {
 
-    protected static final String TAG = SpeechRecognizer.class.getSimpleName();
+    protected static final String TAG = SpeechService.class.getSimpleName();
 
     private final KaldiRecognizer recognizer;
 
-    private final int sampleRate;        
+    private final int sampleRate;
     private final static float BUFFER_SIZE_SECONDS = 0.4f;
     private int bufferSize;
     private final AudioRecord recorder;
@@ -53,17 +52,18 @@ public class SpeechRecognizer {
     private final Collection<RecognitionListener> listeners = new HashSet<RecognitionListener>();
     
     /**
-     * Creates speech recognizer. Recognizer holds the AudioRecord object, so you 
+     * Creates speech service. Service holds the AudioRecord object, so you 
      * need to call {@link release} in order to properly finalize it.
      * 
      * @throws IOException thrown if audio recorder can not be created for some reason.
      */
-    public SpeechRecognizer(Model model) throws IOException {
-        recognizer = new KaldiRecognizer(model, 16000.0f);
-        sampleRate = 16000;
-        bufferSize = Math.round(sampleRate * BUFFER_SIZE_SECONDS);
+    public SpeechService(KaldiRecognizer recognizer, float sampleRate) throws IOException {
+        this.recognizer = recognizer;
+        this.sampleRate = (int)sampleRate;
+
+        bufferSize = Math.round(this.sampleRate * BUFFER_SIZE_SECONDS);
         recorder = new AudioRecord(
-                AudioSource.VOICE_RECOGNITION, sampleRate,
+                AudioSource.VOICE_RECOGNITION, this.sampleRate,
                 AudioFormat.CHANNEL_IN_MONO,
                 AudioFormat.ENCODING_PCM_16BIT, bufferSize * 2);
 
@@ -148,8 +148,7 @@ public class SpeechRecognizer {
     public boolean stop() {
         boolean result = stopRecognizerThread();
         if (result) {
-            Log.i(TAG, "Stop recognition");
-            mainHandler.post(new ResultEvent(recognizer.FinalResult(), true));
+            mainHandler.post(new ResultEvent(recognizer.Result(), true));
         }
         return result;
     }
@@ -162,10 +161,7 @@ public class SpeechRecognizer {
      */
     public boolean cancel() {
         boolean result = stopRecognizerThread();
-        if (result) {
-            Log.i(TAG, "Cancel recognition");
-        }
-
+        recognizer.Result(); // Reset recognizer state
         return result;
     }
     
@@ -175,9 +171,9 @@ public class SpeechRecognizer {
     public void shutdown() {
         recorder.release();
     }
-    
+
     private final class RecognizerThread extends Thread {
-        
+
         private int remainingSamples;
         private int timeoutSamples;
         private final static int NO_TIMEOUT = -1;
@@ -206,8 +202,6 @@ public class SpeechRecognizer {
                 return;
             }
 
-            Log.d(TAG, "Starting decoding");
-
             short[] buffer = new short[bufferSize];
 
             while (!interrupted()

c/Makefile

@@ -0,0 +1,50 @@
+KALDI_ROOT=$(HOME)/travis/kaldi
+
+VOSK_SOURCES= \
+    ../src/kaldi_recognizer.cc \
+    ../src/language_model.cc \
+    ../src/model.cc \
+    ../src/spk_model.cc \
+    ../src/vosk_api.cc
+
+CFLAGS=-g -O2 -DFST_NO_DYNAMIC_LINKING -I../src -I$(KALDI_ROOT)/src -I$(KALDI_ROOT)/tools/openfst/include
+LIBS= \
+	$(KALDI_ROOT)/src/online2/kaldi-online2.a \
+	$(KALDI_ROOT)/src/decoder/kaldi-decoder.a \
+	$(KALDI_ROOT)/src/ivector/kaldi-ivector.a \
+	$(KALDI_ROOT)/src/gmm/kaldi-gmm.a \
+	$(KALDI_ROOT)/src/nnet3/kaldi-nnet3.a \
+	$(KALDI_ROOT)/src/tree/kaldi-tree.a \
+	$(KALDI_ROOT)/src/feat/kaldi-feat.a \
+	$(KALDI_ROOT)/src/lat/kaldi-lat.a \
+	$(KALDI_ROOT)/src/lm/kaldi-lm.a \
+	$(KALDI_ROOT)/src/hmm/kaldi-hmm.a \
+	$(KALDI_ROOT)/src/transform/kaldi-transform.a \
+	$(KALDI_ROOT)/src/cudamatrix/kaldi-cudamatrix.a \
+	$(KALDI_ROOT)/src/matrix/kaldi-matrix.a \
+	$(KALDI_ROOT)/src/fstext/kaldi-fstext.a \
+	$(KALDI_ROOT)/src/util/kaldi-util.a \
+	$(KALDI_ROOT)/src/base/kaldi-base.a \
+	$(KALDI_ROOT)/tools/OpenBLAS/libopenblas.a \
+	$(KALDI_ROOT)/tools/openfst/lib/libfst.a \
+	$(KALDI_ROOT)/tools/openfst/lib/libfstngram.a
+
+all: test_vosk test_vosk_speaker
+
+test_vosk: test_vosk.o libvosk.a
+	g++ $^ -o $@ $(LIBS) -lgfortran -lpthread
+
+test_vosk_speaker: test_vosk_speaker.o libvosk.a
+	g++ $^ -o $@ $(LIBS) -lgfortran -lpthread
+
+libvosk.a: $(VOSK_SOURCES:.cc=.o)
+	ar rcs $@ $^
+
+%.o: %.c
+	g++ $(CFLAGS) -c -o $@ $<
+
+%.o: %.cc
+	g++ -std=c++11 $(CFLAGS) -c -o $@ $<
+
+clean:
+	rm -f *.o *.a test_vosk

c/test_vosk.c

@@ -0,0 +1,28 @@
+#include <vosk_api.h>
+#include <stdio.h>
+
+int main() {
+    FILE *wavin;
+    char buf[3200];
+    int nread, final;
+
+    VoskModel *model = vosk_model_new("model");
+    VoskRecognizer *recognizer = vosk_recognizer_new(model, 16000.0);
+
+    wavin = fopen("test.wav", "rb");
+    fseek(wavin, 44, SEEK_SET);
+    while (!feof(wavin)) {
+         nread = fread(buf, 1, sizeof(buf), wavin);
+         final = vosk_recognizer_accept_waveform(recognizer, buf, nread);
+         if (final) {
+             printf("%s\n", vosk_recognizer_result(recognizer));
+         } else {
+             printf("%s\n", vosk_recognizer_partial_result(recognizer));
+         }
+    }
+    printf("%s\n", vosk_recognizer_final_result(recognizer));
+
+    vosk_recognizer_free(recognizer);
+    vosk_model_free(model);
+    return 0;
+}

c/test_vosk_speaker.c

@@ -0,0 +1,30 @@
+#include <vosk_api.h>
+#include <stdio.h>
+
+int main() {
+    FILE *wavin;
+    char buf[3200];
+    int nread, final;
+
+    VoskModel *model = vosk_model_new("model");
+    VoskSpkModel *spk_model = vosk_spk_model_new("spk-model");
+    VoskRecognizer *recognizer = vosk_recognizer_new_spk(model, spk_model, 16000.0);
+
+    wavin = fopen("test.wav", "rb");
+    fseek(wavin, 44, SEEK_SET);
+    while (!feof(wavin)) {
+         nread = fread(buf, 1, sizeof(buf), wavin);
+         final = vosk_recognizer_accept_waveform(recognizer, buf, nread);
+         if (final) {
+             printf("%s\n", vosk_recognizer_result(recognizer));
+         } else {
+             printf("%s\n", vosk_recognizer_partial_result(recognizer));
+         }
+    }
+    printf("%s\n", vosk_recognizer_final_result(recognizer));
+
+    vosk_recognizer_free(recognizer);
+    vosk_spk_model_free(spk_model);
+    vosk_model_free(model);
+    return 0;
+}

csharp/Makefile

@@ -11,6 +11,7 @@ KALDI_LIBS = \
              ${KALDI_ROOT}/src/tree/kaldi-tree.a \
              ${KALDI_ROOT}/src/feat/kaldi-feat.a \
              ${KALDI_ROOT}/src/lat/kaldi-lat.a \
+             ${KALDI_ROOT}/src/lm/kaldi-lm.a \
              ${KALDI_ROOT}/src/hmm/kaldi-hmm.a \
              ${KALDI_ROOT}/src/transform/kaldi-transform.a \
              ${KALDI_ROOT}/src/cudamatrix/kaldi-cudamatrix.a \
@@ -19,9 +20,12 @@ KALDI_LIBS = \
              ${KALDI_ROOT}/src/util/kaldi-util.a \
              ${KALDI_ROOT}/src/base/kaldi-base.a \
              ${KALDI_ROOT}/tools/openfst/lib/libfst.a \
-             ${KALDI_ROOT}/tools/openfst/lib/libfstngram.a \
-             ${KALDI_ROOT}/tools/OpenBLAS/libopenblas.a \
-             -lgfortran
+             ${KALDI_ROOT}/tools/openfst/lib/libfstngram.a
+
+# On Linux
+MATH_LIBS = ${KALDI_ROOT}/tools/OpenBLAS/libopenblas.a -lgfortran
+# On OSX
+# MATH_LIBS = -framework Accelerate
 
 all: test.exe
 
@@ -29,20 +33,30 @@ test.exe: libkaldiwrap.so test.cs
 	mcs test.cs gen/*.cs
 
 VOSK_SOURCES = \
-	vosk_wrap.cc \
+	vosk_wrap.c \
 	../src/kaldi_recognizer.cc \
-	../src/kaldi_recognizer.h \
+	../src/language_model.cc \
 	../src/model.cc \
-	../src/model.h \
 	../src/spk_model.cc \
+	../src/vosk_api.cc
+
+VOSK_HEADERS = \
+	../src/kaldi_recognizer.h \
+	../src/language_model.h \
+	../src/model.h \
+	../src/vosk_api.h \
 	../src/spk_model.h
 
-libkaldiwrap.so: $(VOSK_SOURCES)
-	$(CXX) -fpermissive $(CFLAGS) $(CPPFLAGS) -shared -o $@ $(VOSK_SOURCES) $(KALDI_LIBS)
+libkaldiwrap.so: $(VOSK_SOURCES) $(VOSK_HEADERS)
+	$(CXX) -fpermissive $(CFLAGS) $(CPPFLAGS) -shared -o $@ $(VOSK_SOURCES) $(KALDI_LIBS) $(MATH_LIBS)
 
-vosk_wrap.cc: ../src/vosk.i
-	swig -csharp -dllimport "libkaldiwrap.so" \
-	    -namespace "Kaldi" -c++ -outdir gen -o vosk_wrap.cc ../src/vosk.i
+vosk_wrap.c: ../src/vosk.i
+	mkdir -p gen
+	swig -csharp -DSWIG_CSHARP_NO_EXCEPTION_HELPER -dllimport "libkaldiwrap" \
+	    -namespace "Kaldi" -outdir gen -o vosk_wrap.c ../src/vosk.i
+
+run: test.exe
+	mono test.exe
 
 clean:
-	$(RM) *.so vosk_wrap.cc *.o gen/*.cs test.exe
+	$(RM) *.so vosk_wrap.c *.o gen/*.cs test.exe

csharp/test.cs

@@ -7,6 +7,7 @@ public class Test
    public static void Main()
    {
 
+        Vosk.SetLogLevel(0);
         Model model = new Model("model");
         KaldiRecognizer rec = new KaldiRecognizer(model, 16000.0f);
 

doc/accuracy.md

@@ -0,0 +1 @@
+See https://alphacephei.com/vosk/accuracy

doc/adaptation.md

@@ -0,0 +1 @@
+See https://alphacephei.com/vosk/adaptation

doc/model.md

@@ -1,42 +0,0 @@
-## Updating the language model
-
-The Kaldi model used in Vosk is compiled from 3 data sources:
-
-  * dictionary
-  * acoustic model
-  * language model
-
-You can rebuild all three with different level of effort, but sometimes you just
-need to adjust the probability of the words to improve the recognition. For
-that it is enough to recompile the language model from the text. To do that
-
-1) Take a text that reflects the speech you want to recognize
-2) Remove punctuation, convert everything to the lowercase, you can do it with a python script
-3) Build openfst and opengrm inside kaldi
-
-```
-export KALDI_ROOT=`pwd`/kaldi
-git clone https://github.com/kaldi-asr/kaldi
-cd kaldi/tools
-make
-# install all required dependencies and repeat `make` if needed
-extras/install_opengrm.sh
-```
-
-4) Now lets build a grammar
-
-```
-export PATH=$KALDI_ROOT/tools/openfst/bin:$PATH
-export LD_LIBRARY_PATH=$KALDI_ROOT/tools/openfst/lib/fst
-cd model
-fstsymbols --save_osymbols=words.txt Gr.fst > /dev/null
-farcompilestrings --fst_type=compact --symbols=words.txt --keep_symbols text.txt | \
-    ngramcount | ngrammake | \
-    fstconvert --fst_type=ngram > Gr.fst
-```
-
-Use created Gr.fst instead of standard one in your model.
-
-For more details see OpenGRM documentation http://www.opengrm.org/twiki/bin/view/GRM/NGramLibrary
-
-You can not introduce new words this way, that is something we will cover later.

doc/models.md

@@ -0,0 +1 @@
+See https://alphacephei.com/vosk/models

ios/README

@@ -0,0 +1,3 @@
+This is a baseline for the vosk-api iOS demo. It requires a build of a
+Vosk-API library, mail contact@alphacephei.com for the details.
+

ios/VoskApiTest.xcodeproj/project.pbxproj

@@ -0,0 +1,477 @@
+// !$*UTF8*$!
+{
+	archiveVersion = 1;
+	classes = {
+	};
+	objectVersion = 46;
+	objects = {
+
+/* Begin PBXBuildFile section */
+		92375222240C550B00DD6076 /* AppDelegate.swift in Sources */ = {isa = PBXBuildFile; fileRef = 92375221240C550B00DD6076 /* AppDelegate.swift */; };
+		92375224240C550B00DD6076 /* ViewController.swift in Sources */ = {isa = PBXBuildFile; fileRef = 92375223240C550B00DD6076 /* ViewController.swift */; };
+		92375227240C550B00DD6076 /* Main.storyboard in Resources */ = {isa = PBXBuildFile; fileRef = 92375225240C550B00DD6076 /* Main.storyboard */; };
+		92375229240C550B00DD6076 /* Assets.xcassets in Resources */ = {isa = PBXBuildFile; fileRef = 92375228240C550B00DD6076 /* Assets.xcassets */; };
+		9237522C240C550B00DD6076 /* LaunchScreen.storyboard in Resources */ = {isa = PBXBuildFile; fileRef = 9237522A240C550B00DD6076 /* LaunchScreen.storyboard */; };
+		92375234240C558900DD6076 /* Vosk.swift in Sources */ = {isa = PBXBuildFile; fileRef = 92375233240C558900DD6076 /* Vosk.swift */; };
+		9237523C240C642000DD6076 /* libkaldiwrap.a in Frameworks */ = {isa = PBXBuildFile; fileRef = 9237523A240C642000DD6076 /* libkaldiwrap.a */; };
+		92375244240C6DAF00DD6076 /* Accelerate.framework in Frameworks */ = {isa = PBXBuildFile; fileRef = 92375243240C6DAF00DD6076 /* Accelerate.framework */; };
+		92375246240C6DC900DD6076 /* libstdc++.tbd in Frameworks */ = {isa = PBXBuildFile; fileRef = 92375245240C6DC900DD6076 /* libstdc++.tbd */; };
+		92375266240C6EFE00DD6076 /* disambig_tid.int in CopyFiles */ = {isa = PBXBuildFile; fileRef = 92375248240C6E3D00DD6076 /* disambig_tid.int */; };
+		92375267240C6EFE00DD6076 /* final.mdl in CopyFiles */ = {isa = PBXBuildFile; fileRef = 92375249240C6E3D00DD6076 /* final.mdl */; };
+		92375268240C6EFE00DD6076 /* Gr.fst in CopyFiles */ = {isa = PBXBuildFile; fileRef = 9237524A240C6E3D00DD6076 /* Gr.fst */; };
+		92375269240C6EFE00DD6076 /* HCLr.fst in CopyFiles */ = {isa = PBXBuildFile; fileRef = 9237524B240C6E3D00DD6076 /* HCLr.fst */; };
+		9237526A240C6EFE00DD6076 /* mfcc.conf in CopyFiles */ = {isa = PBXBuildFile; fileRef = 92375253240C6E3D00DD6076 /* mfcc.conf */; };
+		9237526B240C6EFE00DD6076 /* word_boundary.int in CopyFiles */ = {isa = PBXBuildFile; fileRef = 92375254240C6E3D00DD6076 /* word_boundary.int */; };
+		9237526C240C6EFE00DD6076 /* words.txt in CopyFiles */ = {isa = PBXBuildFile; fileRef = 92375255240C6E3D00DD6076 /* words.txt */; };
+		9237526E240C6F1500DD6076 /* final.dubm in CopyFiles */ = {isa = PBXBuildFile; fileRef = 9237524D240C6E3D00DD6076 /* final.dubm */; };
+		9237526F240C6F1500DD6076 /* final.ie in CopyFiles */ = {isa = PBXBuildFile; fileRef = 9237524E240C6E3D00DD6076 /* final.ie */; };
+		92375270240C6F1500DD6076 /* final.mat in CopyFiles */ = {isa = PBXBuildFile; fileRef = 9237524F240C6E3D00DD6076 /* final.mat */; };
+		92375271240C6F1500DD6076 /* global_cmvn.stats in CopyFiles */ = {isa = PBXBuildFile; fileRef = 92375250240C6E3D00DD6076 /* global_cmvn.stats */; };
+		92375272240C6F1500DD6076 /* online_cmvn.conf in CopyFiles */ = {isa = PBXBuildFile; fileRef = 92375251240C6E3D00DD6076 /* online_cmvn.conf */; };
+		92375273240C6F1500DD6076 /* splice.conf in CopyFiles */ = {isa = PBXBuildFile; fileRef = 92375252240C6E3D00DD6076 /* splice.conf */; };
+		92375274240C6F1E00DD6076 /* 10001-90210-01803.wav in Resources */ = {isa = PBXBuildFile; fileRef = 92375256240C6E3D00DD6076 /* 10001-90210-01803.wav */; };
+/* End PBXBuildFile section */
+
+/* Begin PBXCopyFilesBuildPhase section */
+		92375265240C6ECF00DD6076 /* CopyFiles */ = {
+			isa = PBXCopyFilesBuildPhase;
+			buildActionMask = 2147483647;
+			dstPath = "model-en";
+			dstSubfolderSpec = 7;
+			files = (
+				92375266240C6EFE00DD6076 /* disambig_tid.int in CopyFiles */,
+				92375267240C6EFE00DD6076 /* final.mdl in CopyFiles */,
+				92375268240C6EFE00DD6076 /* Gr.fst in CopyFiles */,
+				92375269240C6EFE00DD6076 /* HCLr.fst in CopyFiles */,
+				9237526A240C6EFE00DD6076 /* mfcc.conf in CopyFiles */,
+				9237526B240C6EFE00DD6076 /* word_boundary.int in CopyFiles */,
+				9237526C240C6EFE00DD6076 /* words.txt in CopyFiles */,
+			);
+			runOnlyForDeploymentPostprocessing = 0;
+		};
+		9237526D240C6F0400DD6076 /* CopyFiles */ = {
+			isa = PBXCopyFilesBuildPhase;
+			buildActionMask = 2147483647;
+			dstPath = "model-en/ivector";
+			dstSubfolderSpec = 7;
+			files = (
+				9237526E240C6F1500DD6076 /* final.dubm in CopyFiles */,
+				9237526F240C6F1500DD6076 /* final.ie in CopyFiles */,
+				92375270240C6F1500DD6076 /* final.mat in CopyFiles */,
+				92375271240C6F1500DD6076 /* global_cmvn.stats in CopyFiles */,
+				92375272240C6F1500DD6076 /* online_cmvn.conf in CopyFiles */,
+				92375273240C6F1500DD6076 /* splice.conf in CopyFiles */,
+			);
+			runOnlyForDeploymentPostprocessing = 0;
+		};
+/* End PBXCopyFilesBuildPhase section */
+
+/* Begin PBXFileReference section */
+		9237521E240C550B00DD6076 /* VoskApiTest.app */ = {isa = PBXFileReference; explicitFileType = wrapper.application; includeInIndex = 0; path = VoskApiTest.app; sourceTree = BUILT_PRODUCTS_DIR; };
+		92375221240C550B00DD6076 /* AppDelegate.swift */ = {isa = PBXFileReference; lastKnownFileType = sourcecode.swift; path = AppDelegate.swift; sourceTree = "<group>"; };
+		92375223240C550B00DD6076 /* ViewController.swift */ = {isa = PBXFileReference; lastKnownFileType = sourcecode.swift; path = ViewController.swift; sourceTree = "<group>"; };
+		92375226240C550B00DD6076 /* Base */ = {isa = PBXFileReference; lastKnownFileType = file.storyboard; name = Base; path = Base.lproj/Main.storyboard; sourceTree = "<group>"; };
+		92375228240C550B00DD6076 /* Assets.xcassets */ = {isa = PBXFileReference; lastKnownFileType = folder.assetcatalog; path = Assets.xcassets; sourceTree = "<group>"; };
+		9237522B240C550B00DD6076 /* Base */ = {isa = PBXFileReference; lastKnownFileType = file.storyboard; name = Base; path = Base.lproj/LaunchScreen.storyboard; sourceTree = "<group>"; };
+		9237522D240C550B00DD6076 /* Info.plist */ = {isa = PBXFileReference; lastKnownFileType = text.plist.xml; path = Info.plist; sourceTree = "<group>"; };
+		92375233240C558900DD6076 /* Vosk.swift */ = {isa = PBXFileReference; fileEncoding = 4; lastKnownFileType = sourcecode.swift; path = Vosk.swift; sourceTree = "<group>"; };
+		9237523A240C642000DD6076 /* libkaldiwrap.a */ = {isa = PBXFileReference; lastKnownFileType = archive.ar; path = libkaldiwrap.a; sourceTree = "<group>"; };
+		92375243240C6DAF00DD6076 /* Accelerate.framework */ = {isa = PBXFileReference; lastKnownFileType = wrapper.framework; name = Accelerate.framework; path = System/Library/Frameworks/Accelerate.framework; sourceTree = SDKROOT; };
+		92375245240C6DC900DD6076 /* libstdc++.tbd */ = {isa = PBXFileReference; lastKnownFileType = "sourcecode.text-based-dylib-definition"; name = "libstdc++.tbd"; path = "usr/lib/libstdc++.tbd"; sourceTree = SDKROOT; };
+		92375248240C6E3D00DD6076 /* disambig_tid.int */ = {isa = PBXFileReference; fileEncoding = 4; lastKnownFileType = text; path = disambig_tid.int; sourceTree = "<group>"; };
+		92375249240C6E3D00DD6076 /* final.mdl */ = {isa = PBXFileReference; lastKnownFileType = file; path = final.mdl; sourceTree = "<group>"; };
+		9237524A240C6E3D00DD6076 /* Gr.fst */ = {isa = PBXFileReference; lastKnownFileType = file; path = Gr.fst; sourceTree = "<group>"; };
+		9237524B240C6E3D00DD6076 /* HCLr.fst */ = {isa = PBXFileReference; lastKnownFileType = file; path = HCLr.fst; sourceTree = "<group>"; };
+		9237524D240C6E3D00DD6076 /* final.dubm */ = {isa = PBXFileReference; lastKnownFileType = file; path = final.dubm; sourceTree = "<group>"; };
+		9237524E240C6E3D00DD6076 /* final.ie */ = {isa = PBXFileReference; lastKnownFileType = file; path = final.ie; sourceTree = "<group>"; };
+		9237524F240C6E3D00DD6076 /* final.mat */ = {isa = PBXFileReference; lastKnownFileType = file; path = final.mat; sourceTree = "<group>"; };
+		92375250240C6E3D00DD6076 /* global_cmvn.stats */ = {isa = PBXFileReference; fileEncoding = 4; lastKnownFileType = text; path = global_cmvn.stats; sourceTree = "<group>"; };
+		92375251240C6E3D00DD6076 /* online_cmvn.conf */ = {isa = PBXFileReference; fileEncoding = 4; lastKnownFileType = text; path = online_cmvn.conf; sourceTree = "<group>"; };
+		92375252240C6E3D00DD6076 /* splice.conf */ = {isa = PBXFileReference; fileEncoding = 4; lastKnownFileType = text; path = splice.conf; sourceTree = "<group>"; };
+		92375253240C6E3D00DD6076 /* mfcc.conf */ = {isa = PBXFileReference; fileEncoding = 4; lastKnownFileType = text; path = mfcc.conf; sourceTree = "<group>"; };
+		92375254240C6E3D00DD6076 /* word_boundary.int */ = {isa = PBXFileReference; fileEncoding = 4; lastKnownFileType = text; path = word_boundary.int; sourceTree = "<group>"; };
+		92375255240C6E3D00DD6076 /* words.txt */ = {isa = PBXFileReference; fileEncoding = 4; lastKnownFileType = text; path = words.txt; sourceTree = "<group>"; };
+		92375256240C6E3D00DD6076 /* 10001-90210-01803.wav */ = {isa = PBXFileReference; lastKnownFileType = audio.wav; path = "10001-90210-01803.wav"; sourceTree = "<group>"; };
+		92AA22AD244CDD1200DA464B /* vosk_api.h */ = {isa = PBXFileReference; fileEncoding = 4; lastKnownFileType = sourcecode.c.h; path = vosk_api.h; sourceTree = "<group>"; };
+		92AA22AE244CDD5200DA464B /* bridging.h */ = {isa = PBXFileReference; fileEncoding = 4; lastKnownFileType = sourcecode.c.h; path = bridging.h; sourceTree = "<group>"; };
+/* End PBXFileReference section */
+
+/* Begin PBXFrameworksBuildPhase section */
+		9237521B240C550B00DD6076 /* Frameworks */ = {
+			isa = PBXFrameworksBuildPhase;
+			buildActionMask = 2147483647;
+			files = (
+				92375246240C6DC900DD6076 /* libstdc++.tbd in Frameworks */,
+				92375244240C6DAF00DD6076 /* Accelerate.framework in Frameworks */,
+				9237523C240C642000DD6076 /* libkaldiwrap.a in Frameworks */,
+			);
+			runOnlyForDeploymentPostprocessing = 0;
+		};
+/* End PBXFrameworksBuildPhase section */
+
+/* Begin PBXGroup section */
+		92375215240C550A00DD6076 = {
+			isa = PBXGroup;
+			children = (
+				92375239240C642000DD6076 /* Vosk */,
+				92375220240C550B00DD6076 /* VoskApiTest */,
+				9237521F240C550B00DD6076 /* Products */,
+				92375242240C6DAF00DD6076 /* Frameworks */,
+			);
+			sourceTree = "<group>";
+		};
+		9237521F240C550B00DD6076 /* Products */ = {
+			isa = PBXGroup;
+			children = (
+				9237521E240C550B00DD6076 /* VoskApiTest.app */,
+			);
+			name = Products;
+			sourceTree = "<group>";
+		};
+		92375220240C550B00DD6076 /* VoskApiTest */ = {
+			isa = PBXGroup;
+			children = (
+				92AA22AE244CDD5200DA464B /* bridging.h */,
+				92375221240C550B00DD6076 /* AppDelegate.swift */,
+				92375223240C550B00DD6076 /* ViewController.swift */,
+				92375225240C550B00DD6076 /* Main.storyboard */,
+				92375228240C550B00DD6076 /* Assets.xcassets */,
+				9237522A240C550B00DD6076 /* LaunchScreen.storyboard */,
+				9237522D240C550B00DD6076 /* Info.plist */,
+				92375233240C558900DD6076 /* Vosk.swift */,
+			);
+			path = VoskApiTest;
+			sourceTree = "<group>";
+		};
+		92375239240C642000DD6076 /* Vosk */ = {
+			isa = PBXGroup;
+			children = (
+				92375247240C6E3D00DD6076 /* model-android */,
+				92375256240C6E3D00DD6076 /* 10001-90210-01803.wav */,
+				92AA22AD244CDD1200DA464B /* vosk_api.h */,
+				9237523A240C642000DD6076 /* libkaldiwrap.a */,
+			);
+			name = Vosk;
+			path = VoskApiTest/Vosk;
+			sourceTree = "<group>";
+		};
+		92375242240C6DAF00DD6076 /* Frameworks */ = {
+			isa = PBXGroup;
+			children = (
+				92375245240C6DC900DD6076 /* libstdc++.tbd */,
+				92375243240C6DAF00DD6076 /* Accelerate.framework */,
+			);
+			name = Frameworks;
+			sourceTree = "<group>";
+		};
+		92375247240C6E3D00DD6076 /* model-android */ = {
+			isa = PBXGroup;
+			children = (
+				92375248240C6E3D00DD6076 /* disambig_tid.int */,
+				92375249240C6E3D00DD6076 /* final.mdl */,
+				9237524A240C6E3D00DD6076 /* Gr.fst */,
+				9237524B240C6E3D00DD6076 /* HCLr.fst */,
+				9237524C240C6E3D00DD6076 /* ivector */,
+				92375253240C6E3D00DD6076 /* mfcc.conf */,
+				92375254240C6E3D00DD6076 /* word_boundary.int */,
+				92375255240C6E3D00DD6076 /* words.txt */,
+			);
+			path = "model-android";
+			sourceTree = "<group>";
+		};
+		9237524C240C6E3D00DD6076 /* ivector */ = {
+			isa = PBXGroup;
+			children = (
+				9237524D240C6E3D00DD6076 /* final.dubm */,
+				9237524E240C6E3D00DD6076 /* final.ie */,
+				9237524F240C6E3D00DD6076 /* final.mat */,
+				92375250240C6E3D00DD6076 /* global_cmvn.stats */,
+				92375251240C6E3D00DD6076 /* online_cmvn.conf */,
+				92375252240C6E3D00DD6076 /* splice.conf */,
+			);
+			path = ivector;
+			sourceTree = "<group>";
+		};
+/* End PBXGroup section */
+
+/* Begin PBXNativeTarget section */
+		9237521D240C550B00DD6076 /* VoskApiTest */ = {
+			isa = PBXNativeTarget;
+			buildConfigurationList = 92375230240C550B00DD6076 /* Build configuration list for PBXNativeTarget "VoskApiTest" */;
+			buildPhases = (
+				9237521A240C550B00DD6076 /* Sources */,
+				9237521B240C550B00DD6076 /* Frameworks */,
+				9237521C240C550B00DD6076 /* Resources */,
+				92375265240C6ECF00DD6076 /* CopyFiles */,
+				9237526D240C6F0400DD6076 /* CopyFiles */,
+			);
+			buildRules = (
+			);
+			dependencies = (
+			);
+			name = VoskApiTest;
+			productName = VoskApiTest;
+			productReference = 9237521E240C550B00DD6076 /* VoskApiTest.app */;
+			productType = "com.apple.product-type.application";
+		};
+/* End PBXNativeTarget section */
+
+/* Begin PBXProject section */
+		92375216240C550A00DD6076 /* Project object */ = {
+			isa = PBXProject;
+			attributes = {
+				LastSwiftUpdateCheck = 0830;
+				LastUpgradeCheck = 0920;
+				ORGANIZATIONNAME = "Alpha Cephei";
+				TargetAttributes = {
+					9237521D240C550B00DD6076 = {
+						CreatedOnToolsVersion = 8.3.2;
+						LastSwiftMigration = 0920;
+						ProvisioningStyle = Automatic;
+					};
+				};
+			};
+			buildConfigurationList = 92375219240C550A00DD6076 /* Build configuration list for PBXProject "VoskApiTest" */;
+			compatibilityVersion = "Xcode 3.2";
+			developmentRegion = English;
+			hasScannedForEncodings = 0;
+			knownRegions = (
+				en,
+				Base,
+			);
+			mainGroup = 92375215240C550A00DD6076;
+			productRefGroup = 9237521F240C550B00DD6076 /* Products */;
+			projectDirPath = "";
+			projectRoot = "";
+			targets = (
+				9237521D240C550B00DD6076 /* VoskApiTest */,
+			);
+		};
+/* End PBXProject section */
+
+/* Begin PBXResourcesBuildPhase section */
+		9237521C240C550B00DD6076 /* Resources */ = {
+			isa = PBXResourcesBuildPhase;
+			buildActionMask = 2147483647;
+			files = (
+				92375274240C6F1E00DD6076 /* 10001-90210-01803.wav in Resources */,
+				9237522C240C550B00DD6076 /* LaunchScreen.storyboard in Resources */,
+				92375229240C550B00DD6076 /* Assets.xcassets in Resources */,
+				92375227240C550B00DD6076 /* Main.storyboard in Resources */,
+			);
+			runOnlyForDeploymentPostprocessing = 0;
+		};
+/* End PBXResourcesBuildPhase section */
+
+/* Begin PBXSourcesBuildPhase section */
+		9237521A240C550B00DD6076 /* Sources */ = {
+			isa = PBXSourcesBuildPhase;
+			buildActionMask = 2147483647;
+			files = (
+				92375224240C550B00DD6076 /* ViewController.swift in Sources */,
+				92375222240C550B00DD6076 /* AppDelegate.swift in Sources */,
+				92375234240C558900DD6076 /* Vosk.swift in Sources */,
+			);
+			runOnlyForDeploymentPostprocessing = 0;
+		};
+/* End PBXSourcesBuildPhase section */
+
+/* Begin PBXVariantGroup section */
+		92375225240C550B00DD6076 /* Main.storyboard */ = {
+			isa = PBXVariantGroup;
+			children = (
+				92375226240C550B00DD6076 /* Base */,
+			);
+			name = Main.storyboard;
+			sourceTree = "<group>";
+		};
+		9237522A240C550B00DD6076 /* LaunchScreen.storyboard */ = {
+			isa = PBXVariantGroup;
+			children = (
+				9237522B240C550B00DD6076 /* Base */,
+			);
+			name = LaunchScreen.storyboard;
+			sourceTree = "<group>";
+		};
+/* End PBXVariantGroup section */
+
+/* Begin XCBuildConfiguration section */
+		9237522E240C550B00DD6076 /* Debug */ = {
+			isa = XCBuildConfiguration;
+			buildSettings = {
+				ALWAYS_SEARCH_USER_PATHS = NO;
+				CLANG_ANALYZER_NONNULL = YES;
+				CLANG_ANALYZER_NUMBER_OBJECT_CONVERSION = YES_AGGRESSIVE;
+				CLANG_CXX_LANGUAGE_STANDARD = "gnu++0x";
+				CLANG_CXX_LIBRARY = "libc++";
+				CLANG_ENABLE_MODULES = YES;
+				CLANG_ENABLE_OBJC_ARC = YES;
+				CLANG_WARN_BLOCK_CAPTURE_AUTORELEASING = YES;
+				CLANG_WARN_BOOL_CONVERSION = YES;
+				CLANG_WARN_COMMA = YES;
+				CLANG_WARN_CONSTANT_CONVERSION = YES;
+				CLANG_WARN_DIRECT_OBJC_ISA_USAGE = YES_ERROR;
+				CLANG_WARN_DOCUMENTATION_COMMENTS = YES;
+				CLANG_WARN_EMPTY_BODY = YES;
+				CLANG_WARN_ENUM_CONVERSION = YES;
+				CLANG_WARN_INFINITE_RECURSION = YES;
+				CLANG_WARN_INT_CONVERSION = YES;
+				CLANG_WARN_NON_LITERAL_NULL_CONVERSION = YES;
+				CLANG_WARN_OBJC_LITERAL_CONVERSION = YES;
+				CLANG_WARN_OBJC_ROOT_CLASS = YES_ERROR;
+				CLANG_WARN_RANGE_LOOP_ANALYSIS = YES;
+				CLANG_WARN_STRICT_PROTOTYPES = YES;
+				CLANG_WARN_SUSPICIOUS_MOVE = YES;
+				CLANG_WARN_UNREACHABLE_CODE = YES;
+				CLANG_WARN__DUPLICATE_METHOD_MATCH = YES;
+				"CODE_SIGN_IDENTITY[sdk=iphoneos*]" = "iPhone Developer";
+				COPY_PHASE_STRIP = NO;
+				DEBUG_INFORMATION_FORMAT = dwarf;
+				ENABLE_STRICT_OBJC_MSGSEND = YES;
+				ENABLE_TESTABILITY = YES;
+				GCC_C_LANGUAGE_STANDARD = gnu99;
+				GCC_DYNAMIC_NO_PIC = NO;
+				GCC_NO_COMMON_BLOCKS = YES;
+				GCC_OPTIMIZATION_LEVEL = 0;
+				GCC_PREPROCESSOR_DEFINITIONS = (
+					"DEBUG=1",
+					"$(inherited)",
+				);
+				GCC_WARN_64_TO_32_BIT_CONVERSION = YES;
+				GCC_WARN_ABOUT_RETURN_TYPE = YES_ERROR;
+				GCC_WARN_UNDECLARED_SELECTOR = YES;
+				GCC_WARN_UNINITIALIZED_AUTOS = YES_AGGRESSIVE;
+				GCC_WARN_UNUSED_FUNCTION = YES;
+				GCC_WARN_UNUSED_VARIABLE = YES;
+				IPHONEOS_DEPLOYMENT_TARGET = 10.3;
+				MTL_ENABLE_DEBUG_INFO = YES;
+				ONLY_ACTIVE_ARCH = YES;
+				SDKROOT = iphoneos;
+				SWIFT_ACTIVE_COMPILATION_CONDITIONS = DEBUG;
+				SWIFT_OBJC_BRIDGING_HEADER = bridging.h;
+				SWIFT_OPTIMIZATION_LEVEL = "-Onone";
+				SWIFT_SWIFT3_OBJC_INFERENCE = Default;
+				TARGETED_DEVICE_FAMILY = "1,2";
+			};
+			name = Debug;
+		};
+		9237522F240C550B00DD6076 /* Release */ = {
+			isa = XCBuildConfiguration;
+			buildSettings = {
+				ALWAYS_SEARCH_USER_PATHS = NO;
+				CLANG_ANALYZER_NONNULL = YES;
+				CLANG_ANALYZER_NUMBER_OBJECT_CONVERSION = YES_AGGRESSIVE;
+				CLANG_CXX_LANGUAGE_STANDARD = "gnu++0x";
+				CLANG_CXX_LIBRARY = "libc++";
+				CLANG_ENABLE_MODULES = YES;
+				CLANG_ENABLE_OBJC_ARC = YES;
+				CLANG_WARN_BLOCK_CAPTURE_AUTORELEASING = YES;
+				CLANG_WARN_BOOL_CONVERSION = YES;
+				CLANG_WARN_COMMA = YES;
+				CLANG_WARN_CONSTANT_CONVERSION = YES;
+				CLANG_WARN_DIRECT_OBJC_ISA_USAGE = YES_ERROR;
+				CLANG_WARN_DOCUMENTATION_COMMENTS = YES;
+				CLANG_WARN_EMPTY_BODY = YES;
+				CLANG_WARN_ENUM_CONVERSION = YES;
+				CLANG_WARN_INFINITE_RECURSION = YES;
+				CLANG_WARN_INT_CONVERSION = YES;
+				CLANG_WARN_NON_LITERAL_NULL_CONVERSION = YES;
+				CLANG_WARN_OBJC_LITERAL_CONVERSION = YES;
+				CLANG_WARN_OBJC_ROOT_CLASS = YES_ERROR;
+				CLANG_WARN_RANGE_LOOP_ANALYSIS = YES;
+				CLANG_WARN_STRICT_PROTOTYPES = YES;
+				CLANG_WARN_SUSPICIOUS_MOVE = YES;
+				CLANG_WARN_UNREACHABLE_CODE = YES;
+				CLANG_WARN__DUPLICATE_METHOD_MATCH = YES;
+				"CODE_SIGN_IDENTITY[sdk=iphoneos*]" = "iPhone Developer";
+				COPY_PHASE_STRIP = NO;
+				DEBUG_INFORMATION_FORMAT = "dwarf-with-dsym";
+				ENABLE_NS_ASSERTIONS = NO;
+				ENABLE_STRICT_OBJC_MSGSEND = YES;
+				GCC_C_LANGUAGE_STANDARD = gnu99;
+				GCC_NO_COMMON_BLOCKS = YES;
+				GCC_WARN_64_TO_32_BIT_CONVERSION = YES;
+				GCC_WARN_ABOUT_RETURN_TYPE = YES_ERROR;
+				GCC_WARN_UNDECLARED_SELECTOR = YES;
+				GCC_WARN_UNINITIALIZED_AUTOS = YES_AGGRESSIVE;
+				GCC_WARN_UNUSED_FUNCTION = YES;
+				GCC_WARN_UNUSED_VARIABLE = YES;
+				IPHONEOS_DEPLOYMENT_TARGET = 10.3;
+				MTL_ENABLE_DEBUG_INFO = NO;
+				SDKROOT = iphoneos;
+				SWIFT_OBJC_BRIDGING_HEADER = bridging.h;
+				SWIFT_OPTIMIZATION_LEVEL = "-Owholemodule";
+				SWIFT_SWIFT3_OBJC_INFERENCE = Default;
+				TARGETED_DEVICE_FAMILY = "1,2";
+				VALIDATE_PRODUCT = YES;
+			};
+			name = Release;
+		};
+		92375231240C550B00DD6076 /* Debug */ = {
+			isa = XCBuildConfiguration;
+			buildSettings = {
+				ASSETCATALOG_COMPILER_APPICON_NAME = AppIcon;
+				CLANG_ENABLE_MODULES = YES;
+				ENABLE_BITCODE = NO;
+				INFOPLIST_FILE = VoskApiTest/Info.plist;
+				LD_RUNPATH_SEARCH_PATHS = "$(inherited) @executable_path/Frameworks";
+				LIBRARY_SEARCH_PATHS = (
+					"$(inherited)",
+					"$(PROJECT_DIR)/VoskApiTest/Vosk",
+				);
+				PRODUCT_BUNDLE_IDENTIFIER = com.alphacephei.VoskApiTest;
+				PRODUCT_NAME = "$(TARGET_NAME)";
+				SWIFT_INSTALL_OBJC_HEADER = YES;
+				SWIFT_OBJC_BRIDGING_HEADER = VoskApiTest/bridging.h;
+				SWIFT_OPTIMIZATION_LEVEL = "-Onone";
+				SWIFT_SWIFT3_OBJC_INFERENCE = Default;
+				SWIFT_VERSION = 4.0;
+			};
+			name = Debug;
+		};
+		92375232240C550B00DD6076 /* Release */ = {
+			isa = XCBuildConfiguration;
+			buildSettings = {
+				ASSETCATALOG_COMPILER_APPICON_NAME = AppIcon;
+				CLANG_ENABLE_MODULES = YES;
+				ENABLE_BITCODE = NO;
+				INFOPLIST_FILE = VoskApiTest/Info.plist;
+				LD_RUNPATH_SEARCH_PATHS = "$(inherited) @executable_path/Frameworks";
+				LIBRARY_SEARCH_PATHS = (
+					"$(inherited)",
+					"$(PROJECT_DIR)/VoskApiTest/Vosk",
+				);
+				PRODUCT_BUNDLE_IDENTIFIER = com.alphacephei.VoskApiTest;
+				PRODUCT_NAME = "$(TARGET_NAME)";
+				SWIFT_INSTALL_OBJC_HEADER = YES;
+				SWIFT_OBJC_BRIDGING_HEADER = VoskApiTest/bridging.h;
+				SWIFT_SWIFT3_OBJC_INFERENCE = Default;
+				SWIFT_VERSION = 4.0;
+			};
+			name = Release;
+		};
+/* End XCBuildConfiguration section */
+
+/* Begin XCConfigurationList section */
+		92375219240C550A00DD6076 /* Build configuration list for PBXProject "VoskApiTest" */ = {
+			isa = XCConfigurationList;
+			buildConfigurations = (
+				9237522E240C550B00DD6076 /* Debug */,
+				9237522F240C550B00DD6076 /* Release */,
+			);
+			defaultConfigurationIsVisible = 0;
+			defaultConfigurationName = Release;
+		};
+		92375230240C550B00DD6076 /* Build configuration list for PBXNativeTarget "VoskApiTest" */ = {
+			isa = XCConfigurationList;
+			buildConfigurations = (
+				92375231240C550B00DD6076 /* Debug */,
+				92375232240C550B00DD6076 /* Release */,
+			);
+			defaultConfigurationIsVisible = 0;
+			defaultConfigurationName = Release;
+		};
+/* End XCConfigurationList section */
+	};
+	rootObject = 92375216240C550A00DD6076 /* Project object */;
+}

ios/VoskApiTest.xcodeproj/project.xcworkspace/contents.xcworkspacedata

@@ -0,0 +1,7 @@
+<?xml version="1.0" encoding="UTF-8"?>
+<Workspace
+   version = "1.0">
+   <FileRef
+      location = "self:VoskApiTest.xcodeproj">
+   </FileRef>
+</Workspace>

ios/VoskApiTest/AppDelegate.swift

@@ -0,0 +1,46 @@
+//
+//  AppDelegate.swift
+//  VoskApiTest
+//
+//  Created by Niсkolay Shmyrev on 01.03.20.
+//  Copyright © 2020 Alpha Cephei. All rights reserved.
+//
+
+import UIKit
+
+@UIApplicationMain
+class AppDelegate: UIResponder, UIApplicationDelegate {
+
+    var window: UIWindow?
+
+
+    func application(_ application: UIApplication, didFinishLaunchingWithOptions launchOptions: [UIApplicationLaunchOptionsKey: Any]?) -> Bool {
+        // Override point for customization after application launch.
+        return true
+    }
+
+    func applicationWillResignActive(_ application: UIApplication) {
+        // Sent when the application is about to move from active to inactive state. This can occur for certain types of temporary interruptions (such as an incoming phone call or SMS message) or when the user quits the application and it begins the transition to the background state.
+        // Use this method to pause ongoing tasks, disable timers, and invalidate graphics rendering callbacks. Games should use this method to pause the game.
+    }
+
+    func applicationDidEnterBackground(_ application: UIApplication) {
+        // Use this method to release shared resources, save user data, invalidate timers, and store enough application state information to restore your application to its current state in case it is terminated later.
+        // If your application supports background execution, this method is called instead of applicationWillTerminate: when the user quits.
+    }
+
+    func applicationWillEnterForeground(_ application: UIApplication) {
+        // Called as part of the transition from the background to the active state; here you can undo many of the changes made on entering the background.
+    }
+
+    func applicationDidBecomeActive(_ application: UIApplication) {
+        // Restart any tasks that were paused (or not yet started) while the application was inactive. If the application was previously in the background, optionally refresh the user interface.
+    }
+
+    func applicationWillTerminate(_ application: UIApplication) {
+        // Called when the application is about to terminate. Save data if appropriate. See also applicationDidEnterBackground:.
+    }
+
+
+}
+

ios/VoskApiTest/Assets.xcassets/AppIcon.appiconset/Contents.json

@@ -0,0 +1,98 @@
+{
+  "images" : [
+    {
+      "idiom" : "iphone",
+      "size" : "20x20",
+      "scale" : "2x"
+    },
+    {
+      "idiom" : "iphone",
+      "size" : "20x20",
+      "scale" : "3x"
+    },
+    {
+      "idiom" : "iphone",
+      "size" : "29x29",
+      "scale" : "2x"
+    },
+    {
+      "idiom" : "iphone",
+      "size" : "29x29",
+      "scale" : "3x"
+    },
+    {
+      "idiom" : "iphone",
+      "size" : "40x40",
+      "scale" : "2x"
+    },
+    {
+      "idiom" : "iphone",
+      "size" : "40x40",
+      "scale" : "3x"
+    },
+    {
+      "idiom" : "iphone",
+      "size" : "60x60",
+      "scale" : "2x"
+    },
+    {
+      "idiom" : "iphone",
+      "size" : "60x60",
+      "scale" : "3x"
+    },
+    {
+      "idiom" : "ipad",
+      "size" : "20x20",
+      "scale" : "1x"
+    },
+    {
+      "idiom" : "ipad",
+      "size" : "20x20",
+      "scale" : "2x"
+    },
+    {
+      "idiom" : "ipad",
+      "size" : "29x29",
+      "scale" : "1x"
+    },
+    {
+      "idiom" : "ipad",
+      "size" : "29x29",
+      "scale" : "2x"
+    },
+    {
+      "idiom" : "ipad",
+      "size" : "40x40",
+      "scale" : "1x"
+    },
+    {
+      "idiom" : "ipad",
+      "size" : "40x40",
+      "scale" : "2x"
+    },
+    {
+      "idiom" : "ipad",
+      "size" : "76x76",
+      "scale" : "1x"
+    },
+    {
+      "idiom" : "ipad",
+      "size" : "76x76",
+      "scale" : "2x"
+    },
+    {
+      "idiom" : "ipad",
+      "size" : "83.5x83.5",
+      "scale" : "2x"
+    },
+    {
+      "idiom" : "ios-marketing",
+      "size" : "1024x1024",
+      "scale" : "1x"
+    }
+  ],
+  "info" : {
+    "version" : 1,
+    "author" : "xcode"
+  }
+}

ios/VoskApiTest/Base.lproj/LaunchScreen.storyboard

@@ -0,0 +1,27 @@
+<?xml version="1.0" encoding="UTF-8" standalone="no"?>
+<document type="com.apple.InterfaceBuilder3.CocoaTouch.Storyboard.XIB" version="3.0" toolsVersion="11134" systemVersion="15F34" targetRuntime="iOS.CocoaTouch" propertyAccessControl="none" useAutolayout="YES" launchScreen="YES" useTraitCollections="YES" colorMatched="YES" initialViewController="01J-lp-oVM">
+    <dependencies>
+        <plugIn identifier="com.apple.InterfaceBuilder.IBCocoaTouchPlugin" version="11106"/>
+        <capability name="documents saved in the Xcode 8 format" minToolsVersion="8.0"/>
+    </dependencies>
+    <scenes>
+        <!--View Controller-->
+        <scene sceneID="EHf-IW-A2E">
+            <objects>
+                <viewController id="01J-lp-oVM" sceneMemberID="viewController">
+                    <layoutGuides>
+                        <viewControllerLayoutGuide type="top" id="Llm-lL-Icb"/>
+                        <viewControllerLayoutGuide type="bottom" id="xb3-aO-Qok"/>
+                    </layoutGuides>
+                    <view key="view" contentMode="scaleToFill" id="Ze5-6b-2t3">
+                        <rect key="frame" x="0.0" y="0.0" width="375" height="667"/>
+                        <autoresizingMask key="autoresizingMask" widthSizable="YES" heightSizable="YES"/>
+                        <color key="backgroundColor" red="1" green="1" blue="1" alpha="1" colorSpace="custom" customColorSpace="sRGB"/>
+                    </view>
+                </viewController>
+                <placeholder placeholderIdentifier="IBFirstResponder" id="iYj-Kq-Ea1" userLabel="First Responder" sceneMemberID="firstResponder"/>
+            </objects>
+            <point key="canvasLocation" x="53" y="375"/>
+        </scene>
+    </scenes>
+</document>

ios/VoskApiTest/Base.lproj/Main.storyboard

@@ -0,0 +1,32 @@
+<?xml version="1.0" encoding="UTF-8"?>
+<document type="com.apple.InterfaceBuilder3.CocoaTouch.Storyboard.XIB" version="3.0" toolsVersion="13771" targetRuntime="iOS.CocoaTouch" propertyAccessControl="none" useAutolayout="YES" useTraitCollections="YES" colorMatched="YES" initialViewController="BYZ-38-t0r">
+    <device id="retina4_7" orientation="portrait">
+        <adaptation id="fullscreen"/>
+    </device>
+    <dependencies>
+        <deployment identifier="iOS"/>
+        <plugIn identifier="com.apple.InterfaceBuilder.IBCocoaTouchPlugin" version="13772"/>
+        <capability name="documents saved in the Xcode 8 format" minToolsVersion="8.0"/>
+    </dependencies>
+    <scenes>
+        <!--View Controller-->
+        <scene sceneID="tne-QT-ifu">
+            <objects>
+                <viewController id="BYZ-38-t0r" customClass="ViewController" customModule="VoskApiTest" customModuleProvider="target" sceneMemberID="viewController">
+                    <textView key="view" clipsSubviews="YES" multipleTouchEnabled="YES" contentMode="scaleToFill" editable="NO" textAlignment="natural" id="CtX-mx-X98">
+                        <rect key="frame" x="0.0" y="0.0" width="375" height="667"/>
+                        <autoresizingMask key="autoresizingMask" flexibleMaxX="YES" flexibleMaxY="YES"/>
+                        <color key="backgroundColor" white="1" alpha="1" colorSpace="calibratedWhite"/>
+                        <fontDescription key="fontDescription" type="system" pointSize="14"/>
+                        <textInputTraits key="textInputTraits" autocapitalizationType="sentences"/>
+                    </textView>
+                    <connections>
+                        <outlet property="mainText" destination="CtX-mx-X98" id="oJy-5J-NKp"/>
+                    </connections>
+                </viewController>
+                <placeholder placeholderIdentifier="IBFirstResponder" id="dkx-z0-nzr" sceneMemberID="firstResponder"/>
+            </objects>
+            <point key="canvasLocation" x="32.799999999999997" y="32.833583208395808"/>
+        </scene>
+    </scenes>
+</document>

ios/VoskApiTest/Info.plist

@@ -0,0 +1,45 @@
+<?xml version="1.0" encoding="UTF-8"?>
+<!DOCTYPE plist PUBLIC "-//Apple//DTD PLIST 1.0//EN" "http://www.apple.com/DTDs/PropertyList-1.0.dtd">
+<plist version="1.0">
+<dict>
+	<key>CFBundleDevelopmentRegion</key>
+	<string>en</string>
+	<key>CFBundleExecutable</key>
+	<string>$(EXECUTABLE_NAME)</string>
+	<key>CFBundleIdentifier</key>
+	<string>$(PRODUCT_BUNDLE_IDENTIFIER)</string>
+	<key>CFBundleInfoDictionaryVersion</key>
+	<string>6.0</string>
+	<key>CFBundleName</key>
+	<string>$(PRODUCT_NAME)</string>
+	<key>CFBundlePackageType</key>
+	<string>APPL</string>
+	<key>CFBundleShortVersionString</key>
+	<string>1.0</string>
+	<key>CFBundleVersion</key>
+	<string>1</string>
+	<key>LSRequiresIPhoneOS</key>
+	<true/>
+	<key>UILaunchStoryboardName</key>
+	<string>LaunchScreen</string>
+	<key>UIMainStoryboardFile</key>
+	<string>Main</string>
+	<key>UIRequiredDeviceCapabilities</key>
+	<array>
+		<string>armv7</string>
+	</array>
+	<key>UISupportedInterfaceOrientations</key>
+	<array>
+		<string>UIInterfaceOrientationPortrait</string>
+		<string>UIInterfaceOrientationLandscapeLeft</string>
+		<string>UIInterfaceOrientationLandscapeRight</string>
+	</array>
+	<key>UISupportedInterfaceOrientations~ipad</key>
+	<array>
+		<string>UIInterfaceOrientationPortrait</string>
+		<string>UIInterfaceOrientationPortraitUpsideDown</string>
+		<string>UIInterfaceOrientationLandscapeLeft</string>
+		<string>UIInterfaceOrientationLandscapeRight</string>
+	</array>
+</dict>
+</plist>

ios/VoskApiTest/ViewController.swift

@@ -0,0 +1,34 @@
+//
+//  ViewController.swift
+//  VoskApiTest
+//
+//  Created by Niсkolay Shmyrev on 01.03.20.
+//  Copyright © 2020 Alpha Cephei. All rights reserved.
+//
+
+import UIKit
+
+class ViewController: UIViewController {
+
+    @IBOutlet var mainText: UITextView!
+
+    override func viewDidLoad() {
+        super.viewDidLoad()
+    
+        DispatchQueue.global(qos: .userInitiated).async {
+            DispatchQueue.main.async {
+                 self.mainText.text =  "Processing file..."
+            }
+            let vosk = Vosk()
+            let res = vosk.recognizeFile()
+            DispatchQueue.main.async {
+                self.mainText.text = res
+            }
+        }
+    }
+
+    override func didReceiveMemoryWarning() {
+        super.didReceiveMemoryWarning()
+    }
+}
+

ios/VoskApiTest/Vosk.swift

@@ -0,0 +1,37 @@
+//
+//  Vosk.swift
+//  VoskApiTest
+//
+//  Created by Niсkolay Shmyrev on 01.03.20.
+//  Copyright © 2020 Alpha Cephei. All rights reserved.
+//
+
+import Foundation
+
+public final class Vosk {
+    
+    func recognizeFile() -> String {
+        var sres = ""
+        if let resourcePath = Bundle.main.resourcePath {
+
+            let modelPath = resourcePath + "/model-en"
+
+            let model = vosk_model_new(modelPath);
+            let recognizer = vosk_recognizer_new(model, 16000.0)
+
+            let audioFile = URL(fileURLWithPath: resourcePath + "/10001-90210-01803.wav")
+            if let data = try? Data(contentsOf: audioFile) {
+                let _ = data.withUnsafeBytes {
+                    vosk_recognizer_accept_waveform(recognizer, $0, Int32(data.count))
+                }
+                let res = vosk_recognizer_final_result(recognizer);
+                sres = String(validatingUTF8: res!)!;
+                print(sres);
+            }
+
+            vosk_recognizer_free(recognizer)
+            vosk_model_free(model)
+        }
+        return sres
+    }
+}

ios/VoskApiTest/Vosk/vosk_api.h

@@ -0,0 +1,210 @@
+// Copyright 2020 Alpha Cephei Inc.
+//
+// Licensed under the Apache License, Version 2.0 (the "License");
+// you may not use this file except in compliance with the License.
+// You may obtain a copy of the License at
+//
+//       http://www.apache.org/licenses/LICENSE-2.0
+//
+// Unless required by applicable law or agreed to in writing, software
+// distributed under the License is distributed on an "AS IS" BASIS,
+// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+// See the License for the specific language governing permissions and
+// limitations under the License.
+
+/* This header contains the C API for Vosk speech recognition system */
+
+#ifndef VOSK_API_H
+#define VOSK_API_H
+
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+/** Model stores all the data required for recognition
+ *  it contains static data and can be shared across processing
+ *  threads. */
+typedef struct VoskModel VoskModel;
+
+
+/** Speaker model is the same as model but contains the data
+ *  for speaker identification. */
+typedef struct VoskSpkModel VoskSpkModel;
+
+
+/** Recognizer object is the main object which processes data.
+ *  Each recognizer usually runs in own thread and takes audio as input.
+ *  Once audio is processed recognizer returns JSON object as a string
+ *  which represent decoded information - words, confidences, times, n-best lists,
+ *  speaker information and so on */
+typedef struct VoskRecognizer VoskRecognizer;
+
+
+/** Loads model data from the file and returns the model object
+ *
+ * @param model_path: the path of the model on the filesystem
+ @ @returns model object */
+VoskModel *vosk_model_new(const char *model_path);
+
+
+/** Releases the model memory
+ *
+ *  The model object is reference-counted so if some recognizer
+ *  depends on this model, model might still stay alive. When
+ *  last recognizer is released, model will be released too. */
+void vosk_model_free(VoskModel *model);
+
+
+/** Loads speaker model data from the file and returns the model object
+ *
+ * @param model_path: the path of the model on the filesystem
+ * @returns model object */
+VoskSpkModel *vosk_spk_model_new(const char *model_path);
+
+
+/** Releases the model memory
+ *
+ *  The model object is reference-counted so if some recognizer
+ *  depends on this model, model might still stay alive. When
+ *  last recognizer is released, model will be released too. */
+void vosk_spk_model_free(VoskSpkModel *model);
+
+/** Creates the recognizer object
+ *
+ *  The recognizers process the speech and return text using shared model data 
+ *  @param sample_rate The sample rate of the audio you going to feed into the recognizer
+ *  @returns recognizer object */
+VoskRecognizer *vosk_recognizer_new(VoskModel *model, float sample_rate);
+
+
+/** Creates the recognizer object with speaker recognition
+ *
+ *  With the speaker recognition mode the recognizer not just recognize
+ *  text but also return speaker vectors one can use for speaker identification
+ *
+ *  @param spk_model speaker model for speaker identification
+ *  @param sample_rate The sample rate of the audio you going to feed into the recognizer
+ *  @returns recognizer object */
+VoskRecognizer *vosk_recognizer_new_spk(VoskModel *model, VoskSpkModel *spk_model, float sample_rate);
+
+
+/** Creates the recognizer object with the grammar
+ *
+ *  Sometimes when you want to improve recognition accuracy and when you don't need
+ *  to recognize large vocabulary you can specify a list of words to recognize. This
+ *  will improve recognizer speed and accuracy but might return [unk] if user said
+ *  something different.
+ *
+ *  Only recognizers with lookahead models support this type of quick configuration.
+ *  Precompiled HCLG graph models are not supported.
+ *
+ *  @param sample_rate The sample rate of the audio you going to feed into the recognizer
+ *  @param grammar The string with the list of words to recognize, for example "one two three four five [unk]"
+ *
+ *  @returns recognizer object */
+VoskRecognizer *vosk_recognizer_new_grm(VoskModel *model, float sample_rate, const char *grammar);
+
+
+/** Accept voice data
+ *
+ *  accept and process new chunk of voice data
+ *
+ *  @param data - audio data in PCM 16-bit mono format
+ *  @param length - length of the audio data
+ *  @returns true if silence is occured and you can retrieve a new utterance with result method */
+int vosk_recognizer_accept_waveform(VoskRecognizer *recognizer, const char *data, int length);
+
+
+/** Same as above but the version with the short data for language bindings where you have
+ *  audio as array of shorts */
+int vosk_recognizer_accept_waveform_s(VoskRecognizer *recognizer, const short *data, int length);
+
+
+/** Same as above but the version with the float data for language bindings where you have
+ *  audio as array of floats */
+int vosk_recognizer_accept_waveform_f(VoskRecognizer *recognizer, const float *data, int length);
+
+
+/** Returns speech recognition result
+ *
+ * @returns the result in JSON format which contains decoded line, decoded
+ *          words, times in seconds and confidences. You can parse this result
+ *          with any json parser
+ *
+ * <pre>
+ * {
+ *   "result" : [{
+ *       "conf" : 1.000000,
+ *       "end" : 1.110000,
+ *       "start" : 0.870000,
+ *       "word" : "what"
+ *     }, {
+ *       "conf" : 1.000000,
+ *       "end" : 1.530000,
+ *       "start" : 1.110000,
+ *       "word" : "zero"
+ *     }, {
+ *       "conf" : 1.000000,
+ *       "end" : 1.950000,
+ *       "start" : 1.530000,
+ *       "word" : "zero"
+ *     }, {
+ *       "conf" : 1.000000,
+ *       "end" : 2.340000,
+ *       "start" : 1.950000,
+ *       "word" : "zero"
+ *     }, {
+ *       "conf" : 1.000000,
+ *      "end" : 2.610000,
+ *       "start" : 2.340000,
+ *       "word" : "one"
+ *     }],
+ *   "text" : "what zero zero zero one"
+ *  }
+ * </pre>
+ */
+const char *vosk_recognizer_result(VoskRecognizer *recognizer);
+
+
+/** Returns partial speech recognition
+ *
+ * @returns partial speech recognition text which is not yet finalized.
+ *          result may change as recognizer process more data.
+ *
+ * <pre>
+ * {
+ *  "partial" : "cyril one eight zero"
+ * }
+ * </pre>
+ */
+const char *vosk_recognizer_partial_result(VoskRecognizer *recognizer);
+
+
+/** Returns speech recognition result. Same as result, but doesn't wait for silence
+ *  You usually call it in the end of the stream to get final bits of audio. It
+ *  flushes the feature pipeline, so all remaining audio chunks got processed.
+ *
+ *  @returns speech result in JSON format.
+ */
+const char *vosk_recognizer_final_result(VoskRecognizer *recognizer);
+
+
+/** Releases recognizer object
+ *
+ *  Underlying model is also unreferenced and if needed released */
+void vosk_recognizer_free(VoskRecognizer *recognizer);
+
+/** Set log level for Kaldi messages
+ *
+ *  @param log_level the level
+ *     0 - default value to print info and error messages but no debug
+ *     less than 0 - don't print info messages
+ *     greather than 0 - more verbose mode
+ */
+void vosk_set_log_level(int log_level);
+
+#ifdef __cplusplus
+}
+#endif
+
+#endif /* VOSK_API_H */

ios/VoskApiTest/bridging.h

@@ -0,0 +1 @@
+#import "Vosk/vosk_api.h"

java/Makefile

@@ -11,6 +11,7 @@ KALDI_LIBS = \
              ${KALDI_ROOT}/src/tree/kaldi-tree.a \
              ${KALDI_ROOT}/src/feat/kaldi-feat.a \
              ${KALDI_ROOT}/src/lat/kaldi-lat.a \
+             ${KALDI_ROOT}/src/lm/kaldi-lm.a \
              ${KALDI_ROOT}/src/hmm/kaldi-hmm.a \
              ${KALDI_ROOT}/src/transform/kaldi-transform.a \
              ${KALDI_ROOT}/src/cudamatrix/kaldi-cudamatrix.a \
@@ -29,17 +30,21 @@ VOSK_SOURCES = \
 	vosk_wrap.cc \
 	../src/kaldi_recognizer.cc \
 	../src/kaldi_recognizer.h \
+	../src/language_model.cc \
+	../src/language_model.h \
 	../src/model.cc \
 	../src/model.h \
 	../src/spk_model.cc \
-	../src/spk_model.h
+	../src/spk_model.h \
+	../src/vosk_api.cc \
+	../src/vosk_api.h
 
 libvosk_jni.so: $(VOSK_SOURCES)
 	$(CXX) -shared -o $@ $(CPPFLAGS) $(CFLAGS) $(VOSK_SOURCES) $(KALDI_LIBS)
 
 vosk_wrap.cc: ../src/vosk.i
 	mkdir -p org/kaldi
-	swig -I../src -c++ \
+	swig -c++ -I../src \
 		-java -package org.kaldi \
 		-outdir org/kaldi -o $@ $<
 
@@ -47,11 +52,16 @@ clean:
 	$(RM) *.so *_wrap.cc *_wrap.o test/*.class
 	$(RM) -r org model-en
 
-model-en:
-	wget https://github.com/alphacep/kaldi-android-demo/releases/download/2020-01/alphacep-model-android-en-us-0.3.tar.gz
-	tar xf alphacep-model-android-en-us-0.3.tar.gz && rm alphacep-model-android-en-us-0.3.tar.gz
-	mv alphacep-model-android-en-us-0.3 model-en
+model:
+	wget https://alphacephei.com/kaldi/models/vosk-model-small-en-us-0.3.zip
+	unzip vosk-model-small-en-us-0.3.zip && rm vosk-model-small-en-us-0.3.zip
+	mv vosk-model-small-en-us-0.3 model
 
-run: model-en
+model-spk:
+	wget https://alphacephei.com/kaldi/models/vosk-model-spk-0.3.zip
+	unzip vosk-model-spk-0.3.zip && rm vosk-model-spk-0.3.zip
+	mv vosk-model-spk-0.3 model-spk
+
+run: model model-spk
 	javac test/*.java org/kaldi/*.java
 	java -Djava.library.path=. -cp . test.DecoderTest

java/README.md

@@ -0,0 +1,19 @@
+Java API sample
+
+Doesn't work on Windows or Mac yet, help to prepare the packaged jars is welcome.
+
+For now to try it:
+
+On Linux you can do
+
+   1. Build recent kaldi
+   1. `git clone https://github.com/alphacep/vosk-api`
+   1. `cd vosk-api/java`
+   1. `export KALDI_ROOT=<KALDI_ROOT>`
+   1. `export JAVA_HOME=<JAVA_HOME>`
+   1. `make`
+   1. `make run`
+
+For details of the code you can check:
+
+https://github.com/alphacep/vosk-api/blob/master/java/test/DecoderTest.java

java/test/DecoderTest.java

@@ -11,22 +11,22 @@ import java.nio.*;
 import org.kaldi.KaldiRecognizer;
 import org.kaldi.Model;
 import org.kaldi.SpkModel;
+import org.kaldi.Vosk;
 
 public class DecoderTest {
-    static {
-        System.loadLibrary("vosk_jni");
-    }
 
     public static void main(String args[]) throws IOException {
+        Vosk.SetLogLevel(-10);
+
         FileInputStream ais = new FileInputStream(new File("../python/example/test.wav"));
-        Model model = new Model("model-en");
+        Model model = new Model("model");
         SpkModel spkModel = new SpkModel("model-spk");
         KaldiRecognizer rec = new KaldiRecognizer(model, spkModel, 16000.0f);
 
         int nbytes;
         byte[] b = new byte[4096];
         while ((nbytes = ais.read(b)) >= 0) {
-            if (rec.AcceptWaveform(b, nbytes)) {
+            if (rec.AcceptWaveform(b)) {
                 System.out.println(rec.Result());
             } else {
                 System.out.println(rec.PartialResult());

nodejs/.npmignore

@@ -0,0 +1,3 @@
+build
+node_modules
+vosk_wrap.cc

nodejs/README.md

@@ -0,0 +1,20 @@
+Installation requires vosk-api checkout, it doesn't yet work with `npm
+install vosk`. We have to figure out how to properly distribute native
+modules for Vosk.
+
+The build tested with node-0.10.15, node-0.12 is not yet supported by swig.
+
+Still, you need swig of newest version 4.0.1
+
+Build like this
+
+```
+npm install --kaldi_root=/home/user/kaldi
+```
+
+Then test with
+
+```
+cd example
+node test.js
+```

nodejs/binding.gyp

@@ -0,0 +1,80 @@
+{
+  'targets': [
+    {
+      'target_name': 'vosk',
+      'sources': [
+         '../src/kaldi_recognizer.cc',
+         '../src/model.cc',
+         '../src/language_model.cc',
+         '../src/spk_model.cc',
+         '../src/vosk_api.cc',
+         'vosk_wrap.cc',
+      ],
+      'cflags': [
+          '-std=c++11',
+          '-DFST_NO_DYNAMIC_LINKING',
+          '-Wno-deprecated-declarations',
+          '-Wno-sign-compare',
+          '-Wno-unused-local-typedefs',
+          '-Wno-ignored-quaifiers',
+          '-Wno-extra',
+      ],
+      'cflags_cc!' : [
+          '-fno-rtti',
+          '-fno-exceptions',
+      ],
+      'conditions': [
+          ['OS == "mac"', {
+              'xcode_settings': {
+                  'GCC_ENABLE_CPP_EXCEPTIONS': 'YES',
+                  'GCC_ENABLE_CPP_RTTI': 'YES',
+                  'CLANG_CXX_LANGUAGE_STANDARD': 'c++11'
+              }
+          }]
+      ],
+      'actions': [
+        {
+          'action_name': 'swig',
+          'inputs': [
+            '../src/vosk.i',
+          ],
+          'outputs': [
+            'vosk_wrap.cc',
+          ],
+          'action': ['swig', '-c++', '-javascript', '-o', 'vosk_wrap.cc', '-v8', '-DV8_MAJOR_VERSION=10', '../src/vosk.i']
+        },
+      ],
+      'include_dirs': [
+         '<@(kaldi_root)/src',
+         '<@(kaldi_root)/tools/openfst/include',
+         '../src',
+      ],
+      'link_settings': {
+         'libraries': [
+             '<@(kaldi_root)/src/online2/kaldi-online2.a',
+             '<@(kaldi_root)/src/decoder/kaldi-decoder.a',
+             '<@(kaldi_root)/src/ivector/kaldi-ivector.a',
+             '<@(kaldi_root)/src/gmm/kaldi-gmm.a',
+             '<@(kaldi_root)/src/nnet3/kaldi-nnet3.a',
+             '<@(kaldi_root)/src/tree/kaldi-tree.a',
+             '<@(kaldi_root)/src/feat/kaldi-feat.a',
+             '<@(kaldi_root)/src/lat/kaldi-lat.a',
+             '<@(kaldi_root)/src/lm/kaldi-lm.a',
+             '<@(kaldi_root)/src/hmm/kaldi-hmm.a',
+             '<@(kaldi_root)/src/transform/kaldi-transform.a',
+             '<@(kaldi_root)/src/cudamatrix/kaldi-cudamatrix.a',
+             '<@(kaldi_root)/src/matrix/kaldi-matrix.a',
+             '<@(kaldi_root)/src/fstext/kaldi-fstext.a',
+             '<@(kaldi_root)/src/util/kaldi-util.a',
+             '<@(kaldi_root)/src/base/kaldi-base.a',
+             '<@(kaldi_root)/tools/openfst/lib/libfst.a',
+             '<@(kaldi_root)/tools/openfst/lib/libfstngram.a',
+             '<@(kaldi_root)/tools/OpenBLAS/libopenblas.a',
+          ],
+         'library_dirs': [
+               '/usr/lib',
+          ],
+      },
+    }
+  ]
+}

nodejs/example/test-speaker.js

@@ -0,0 +1,90 @@
+const wav = require('wav')
+const fs = require('fs')
+
+const {Readable} = require('stream')
+const {Model, KaldiRecognizer, SpkModel} = require('..')
+
+try {
+    fs.accessSync('model', fs.constants.R_OK)
+} catch (err) {
+    console.error("Please download the model from https://github.com/alphacep/kaldi-android-demo/releases and unpack as 'model' in the current folder.")
+    process.exit(1)
+}
+
+try {
+    fs.accessSync('model-spk', fs.constants.R_OK)
+} catch (err) {
+    console.error("Please download the speaker model from https://github.com/alphacep/kaldi-android-demo/releases and unpack as 'model-spk' in the current folder.")
+    process.exit(1)
+}
+
+const wfStream = fs.createReadStream('test.wav', { highWaterMark: 4096 })
+const wfReader = new wav.Reader()
+
+const model = new Model('model')
+const spkModel = new SpkModel('model-spk')
+
+const spk_sig = [-0.56648, 0.030579, 1.730239, 0.239899, -1.194183,
+0.251954, 0.540388, -0.971872, -0.020963, 0.085036, 0.563973, -0.019682,
+-0.597381, 1.094719, -0.566738, 0.29819, 0.171165, 0.370341, -0.033539,
+-0.09757, 1.228286, 0.485949, 0.427826, 0.147762, -0.015112, 0.599513,
+-2.040655, -0.490882, 0.440161, -0.072991, 0.835955, -0.496124, 0.952978,
+0.85356, -1.096116, 0.107764, -0.385486, 1.410305, 0.609147, -0.457014,
+-1.542864, 0.343669, 0.171913, -0.627281, -1.281781, -1.134276,
+-0.639895, 1.190183, -0.700537, 1.063457, 0.206946, 0.342198, -1.165625,
+1.475955, -0.089007, -2.555155, 0.551438, -0.212736, 1.025625, -1.631965,
+-0.716256, -1.295995, 1.554956, -1.866009, -1.010782, -1.43231, 0.109027,
+2.123925, 1.703283, -0.784997, 2.730568, 0.755113, 0.0617, 0.128955,
+-0.054047, 1.359119, -0.611666, -1.105754, -0.631353, 0.052109, 0.729386,
+-0.769876, 1.250235, -1.463298, 0.648176, -0.73239, -0.385239,
+-1.661856, 0.602106, -0.45567, -1.438431, -0.836673, 0.033557, 0.373597,
+-1.343341, -0.181095, 0.237287, -0.522005, -1.722836, 0.932333,
+-0.092861, -0.219254, 0.476182, 1.033803, -1.633563, -0.874341, 1.039064,
+-1.758573, -0.838422, -0.324336, -0.924634, 1.962594, 2.152814, 1.2521,
+-0.46172, -1.50271, 1.685691, 0.403097, -0.819042, 0.866403, -0.591716,
+-0.578645, -0.553839, 0.381861, -1.051647, -1.477578, 0.524005, 0.925245]
+
+function dotp(x, y) {
+    function dotp_sum(a, b) {
+        return a + b
+    }
+    function dotp_times(a, i) {
+        return x[i] * y[i]
+    }
+    return x.map(dotp_times).reduce(dotp_sum, 0)
+}
+
+function cosineSimilarity(A, B) {
+    var similarity =
+        dotp(A, B) / (Math.sqrt(dotp(A, A)) * Math.sqrt(dotp(B, B)))
+        return similarity
+}
+
+function cosine_dist(x, y) {
+    return 1 - cosineSimilarity(x, y)
+}
+
+wfReader.on('format', async ({ audioFormat, sampleRate, channels }) => {
+    if (audioFormat != 1 || channels != 1) {
+        console.error('Audio file must be WAV format mono PCM.')
+        process.exit(1)
+    }
+    const rec = new KaldiRecognizer(model, spkModel, sampleRate)
+    for await (const data of new Readable().wrap(wfReader)) {
+        const endOfSpeech = await rec.AcceptWaveform(data)
+        if (endOfSpeech) {
+            res = await JSON.parse(rec.Result());
+            console.log(res)
+            console.log('X-vector:', JSON.stringify(res['spk']))
+            console.log('Speaker distance:', cosine_dist(spk_sig, res['spk']))
+        } else {
+            console.log(await rec.PartialResult())
+        }
+    }
+    res = await JSON.parse(rec.FinalResult());
+    console.log(res)
+    console.log('X-vector:', JSON.stringify(res['spk']))
+    console.log('Speaker distance:', cosine_dist(spk_sig, res['spk']))
+})
+
+wfStream.pipe(wfReader)

nodejs/example/test.js

@@ -0,0 +1,38 @@
+#!/usr/bin/env node
+
+const fs = require("fs");
+const { Readable } = require("stream");
+const wav = require("wav");
+
+const { Model, KaldiRecognizer } = require("..");
+
+try {
+    fs.accessSync("model", fs.constants.R_OK);
+} catch(err) {
+    console.error("Please download the model from https://github.com/alphacep/kaldi-android-demo/releases and unpack as 'model' in the current folder.");
+    process.exit(1);
+}
+
+const wfStream = fs.createReadStream("test.wav", {'highWaterMark': 4096});
+const wfReader = new wav.Reader();
+
+const model = new Model("model");
+
+wfReader.on('format', async ({ audioFormat, sampleRate, channels }) => {
+    if (audioFormat != 1 || channels != 1) {
+        console.error("Audio file must be WAV format mono PCM.");
+        process.exit(1);
+    }
+    const rec = new KaldiRecognizer(model, sampleRate);
+    for await (const data of new Readable().wrap(wfReader)) {
+        const result = await rec.AcceptWaveform(data);
+        if (result != 0) {
+           console.log(await rec.Result());
+        } else {
+           console.log(await rec.PartialResult());
+        }
+    }
+    console.log(await rec.FinalResult());
+});
+
+wfStream.pipe(wfReader);

nodejs/index.js

@@ -1,3 +1,3 @@
-exports.printMsg = function() {
-  console.log("This is a message from the Vosk package");
-}
+const voskNativeModule = require('./build/Release/vosk.node');
+
+module.exports = voskNativeModule;

nodejs/package.json

@@ -1,7 +1,7 @@
 {
   "name": "vosk",
-  "version": "0.1.0",
-  "description": "Node binding for continuous voice recoginition through pocketsphinx.",
+  "version": "0.3.15",
+  "description": "Node binding for continuous offline voice recoginition with Vosk library.",
   "repository": {
     "type": "git",
     "url": "git://github.com/alphacep/vosk-api.git"
@@ -13,6 +13,12 @@
     "voice"
   ],
   "author": "Alpha Cephei Inc.",
-  "license": "Apache 2.0",
-  "engines": { "node" : ">= 12.x.x" }
+  "license": "Apache-2.0",
+  "engines": {
+    "node": ">= 10.x.x"
+  },
+  "dependencies": {
+    "node-gyp": "^5.1.1",
+    "wav": "^1.0.2"
+  }
 }

python/CMakeLists.txt

@@ -1,50 +0,0 @@
-cmake_minimum_required(VERSION 3.12.0)
-project(vosk)
-
-set(TOP_SRCDIR "${CMAKE_SOURCE_DIR}/..")
-if("x$ENV{WHEEL_FLAGS}" STREQUAL "x")
-    find_package (Python COMPONENTS Interpreter Development)
-else()
-    # docker case
-    set(Python_INCLUDE_DIR "")
-    set(Python_LIBRARY "")
-    set(TOP_SRCDIR "/io")
-endif()
-
-set(KALDI_ROOT "$ENV{KALDI_ROOT}")
-set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -std=c++11 -O3 -DFST_NO_DYNAMIC_LINKING")
-include_directories("${TOP_SRCDIR}/src" "${KALDI_ROOT}/src" "${KALDI_ROOT}/tools/openfst/include" ${Python_INCLUDE_DIR})
-set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} $ENV{WHEEL_FLAGS}")
-
-find_package(SWIG REQUIRED)
-include(${SWIG_USE_FILE})
-
-set_property(SOURCE "${TOP_SRCDIR}/src/vosk.i" PROPERTY CPLUSPLUS ON)
-swig_add_library(vosk TYPE SHARED LANGUAGE Python OUTPUT_DIR "${CMAKE_LIBRARY_OUTPUT_DIRECTORY}" OUTFILE_DIR "."
-    SOURCES "${TOP_SRCDIR}/src/kaldi_recognizer.cc" 
-            "${TOP_SRCDIR}/src/spk_model.cc"
-            "${TOP_SRCDIR}/src/model.cc"
-            "${TOP_SRCDIR}/src/vosk.i")
-
-swig_link_libraries(vosk
-             ${KALDI_ROOT}/src/online2/kaldi-online2.a
-             ${KALDI_ROOT}/src/decoder/kaldi-decoder.a
-             ${KALDI_ROOT}/src/ivector/kaldi-ivector.a
-             ${KALDI_ROOT}/src/gmm/kaldi-gmm.a
-             ${KALDI_ROOT}/src/nnet3/kaldi-nnet3.a
-             ${KALDI_ROOT}/src/tree/kaldi-tree.a
-             ${KALDI_ROOT}/src/feat/kaldi-feat.a
-             ${KALDI_ROOT}/src/lat/kaldi-lat.a
-             ${KALDI_ROOT}/src/hmm/kaldi-hmm.a
-             ${KALDI_ROOT}/src/transform/kaldi-transform.a
-             ${KALDI_ROOT}/src/cudamatrix/kaldi-cudamatrix.a
-             ${KALDI_ROOT}/src/matrix/kaldi-matrix.a
-             ${KALDI_ROOT}/src/fstext/kaldi-fstext.a
-             ${KALDI_ROOT}/src/util/kaldi-util.a
-             ${KALDI_ROOT}/src/base/kaldi-base.a
-             ${KALDI_ROOT}/tools/openfst/lib/libfst.a
-             ${KALDI_ROOT}/tools/openfst/lib/libfstngram.a
-             ${KALDI_ROOT}/tools/OpenBLAS/libopenblas.a
-             -lgfortran)
-
-set_target_properties(_vosk PROPERTIES LINK_FLAGS_RELEASE -s)

python/README.md

@@ -1,3 +1,23 @@
-Python module for vosk-api
+This is a Python module for Vosk.
 
-See for details https://github.com/alphacep/vosk-api
+Vosk is an offline open source speech recognition toolkit. It enables
+speech recognition models for 16 languages and dialects - English, Indian
+English, German, French, Spanish, Portuguese, Chinese, Russian, Turkish,
+Vietnamese, Italian, Dutch, Catalan, Arabic, Greek, Farsi.
+
+Vosk models are small (50 Mb) but provide continuous large vocabulary
+transcription, zero-latency response with streaming API, reconfigurable
+vocabulary and speaker identification.
+
+Vosk supplies speech recognition for chatbots, smart home appliances,
+virtual assistants. It can also create subtitles for movies,
+transcription for lectures and interviews.
+
+Vosk scales from small devices like Raspberry Pi or Android smartphone to
+big clusters.
+
+# Documentation
+
+For installation instructions, examples and documentation visit [Vosk
+Website](https://alphacephei.com/vosk). See also our project on
+[Github](https://github.com/alphacep/vosk-api).

python/cmake.py

@@ -1,76 +0,0 @@
-# From https://github.com/raydouglass/cmake_setuptools
-
-import os
-import subprocess
-import shutil
-import sys
-from setuptools import Extension
-from setuptools.command.build_ext import build_ext
-from setuptools.command.build_py import build_py
-
-CMAKE_EXE = os.environ.get('CMAKE_EXE', shutil.which('cmake'))
-
-
-def check_for_cmake():
-    if not CMAKE_EXE:
-        print('cmake executable not found. '
-              'Set CMAKE_EXE environment or update your path')
-        sys.exit(1)
-
-
-class CMakeExtension(Extension):
-    """
-    setuptools.Extension for cmake
-    """
-
-    def __init__(self, name, pkg_name, sourcedir=''):
-        check_for_cmake()
-        Extension.__init__(self, name, sources=[])
-        self.sourcedir = os.path.abspath(sourcedir)
-        self.pkg_name = pkg_name
-
-
-class CMakeBuildExt(build_ext):
-    """
-    setuptools build_exit which builds using cmake & make
-    You can add cmake args with the CMAKE_COMMON_VARIABLES environment variable
-    """
-
-    def build_extension(self, ext):
-        check_for_cmake()
-        if isinstance(ext, CMakeExtension):
-            output_dir = os.path.abspath(
-                os.path.dirname(self.get_ext_fullpath(ext.pkg_name + "/" + ext.name)))
-
-            build_type = 'Debug' if self.debug else 'Release'
-            cmake_args = [CMAKE_EXE,
-                          ext.sourcedir,
-                          '-Wno-dev',
-                          '-DCMAKE_LIBRARY_OUTPUT_DIRECTORY=' + output_dir,
-                          '-DCMAKE_BUILD_TYPE=' + build_type]
-            cmake_args.extend(
-                [x for x in
-                 os.environ.get('CMAKE_COMMON_VARIABLES', '').split(' ')
-                 if x])
-
-            env = os.environ.copy()
-            if not os.path.exists(self.build_temp):
-                os.makedirs(self.build_temp)
-            subprocess.check_call(cmake_args,
-                                  cwd=self.build_temp,
-                                  env=env)
-            subprocess.check_call(['make', 'VERBOSE=1', ext.name],
-                                  cwd=self.build_temp,
-                                  env=env)
-            print()
-        else:
-            super().build_extension(ext)
-
-
-
-class CMakeBuildExtFirst(build_py):
-    def run(self):
-        self.run_command("build_ext")
-        return super().run()
-
-__all__ = ['CMakeBuildExt', 'CMakeExtension', 'CMakeBuildExtFirst']

python/example/test_empty.py

@@ -1,10 +1,10 @@
-#!/usr/bin/python3
+#!/usr/bin/env python3
 
 from vosk import Model, KaldiRecognizer
 import sys
 import json
 
-model = Model("model-en")
+model = Model("model")
 rec = KaldiRecognizer(model, 8000)
 
 res = json.loads(rec.FinalResult())

python/example/test_ffmpeg.py

@@ -0,0 +1,33 @@
+#!/usr/bin/env python3
+
+from vosk import Model, KaldiRecognizer, SetLogLevel
+import sys
+import os
+import wave
+import subprocess
+
+SetLogLevel(0)
+
+if not os.path.exists("model"):
+    print ("Please download the model from https://alphacephei.com/vosk/models and unpack as 'model' in the current folder.")
+    exit (1)
+
+sample_rate=16000
+model = Model("model")
+rec = KaldiRecognizer(model, sample_rate)
+
+process = subprocess.Popen(['ffmpeg', '-loglevel', 'quiet', '-i',
+                            sys.argv[1],
+                            '-ar', str(sample_rate) , '-ac', '1', '-f', 's16le', '-'],
+                            stdout=subprocess.PIPE)
+
+while True:
+    data = process.stdout.read(4000)
+    if len(data) == 0:
+        break
+    if rec.AcceptWaveform(data):
+        print(rec.Result())
+    else:
+        print(rec.PartialResult())
+
+print(rec.FinalResult())

python/example/test_microphone.py

@@ -1,23 +1,23 @@
-#!/usr/bin/python3
+#!/usr/bin/env python3
 
 from vosk import Model, KaldiRecognizer
 import os
 
-if not os.path.exists("model-en"):
-    print ("Please download the model from https://github.com/alphacep/kaldi-android-demo/releases and unpack as 'model' in the current folder.")
+if not os.path.exists("model"):
+    print ("Please download the model from https://alphacephei.com/vosk/models and unpack as 'model' in the current folder.")
     exit (1)
 
 import pyaudio
 
+model = Model("model")
+rec = KaldiRecognizer(model, 16000)
+
 p = pyaudio.PyAudio()
 stream = p.open(format=pyaudio.paInt16, channels=1, rate=16000, input=True, frames_per_buffer=8000)
 stream.start_stream()
 
-model = Model("model-en")
-rec = KaldiRecognizer(model, 16000)
-
 while True:
-    data = stream.read(2000)
+    data = stream.read(4000)
     if len(data) == 0:
         break
     if rec.AcceptWaveform(data):

python/example/test_simple.py

@@ -1,12 +1,14 @@
-#!/usr/bin/python3
+#!/usr/bin/env python3
 
-from vosk import Model, KaldiRecognizer
+from vosk import Model, KaldiRecognizer, SetLogLevel
 import sys
 import os
 import wave
 
-if not os.path.exists("model-en"):
-    print ("Please download the model from https://github.com/alphacep/kaldi-android-demo/releases and unpack as 'model-en' in the current folder.")
+SetLogLevel(0)
+
+if not os.path.exists("model"):
+    print ("Please download the model from https://alphacephei.com/vosk/models and unpack as 'model' in the current folder.")
     exit (1)
 
 wf = wave.open(sys.argv[1], "rb")
@@ -14,11 +16,11 @@ if wf.getnchannels() != 1 or wf.getsampwidth() != 2 or wf.getcomptype() != "NONE
     print ("Audio file must be WAV format mono PCM.")
     exit (1)
 
-model = Model("model-en")
+model = Model("model")
 rec = KaldiRecognizer(model, wf.getframerate())
 
 while True:
-    data = wf.readframes(1000)
+    data = wf.readframes(4000)
     if len(data) == 0:
         break
     if rec.AcceptWaveform(data):

python/example/test_speaker.py

@@ -1,4 +1,4 @@
-#!/usr/bin/python3
+#!/usr/bin/env python3
 
 from vosk import Model, KaldiRecognizer, SpkModel
 import sys
@@ -7,15 +7,15 @@ import json
 import os
 import numpy as np
 
-model_path = "model-en"
+model_path = "model"
 spk_model_path = "model-spk"
 
 if not os.path.exists(model_path):
-    print ("Please download the model from https://github.com/alphacep/kaldi-android-demo/releases and unpack as {} in the current folder.".format(model_path))
+    print ("Please download the model from https://alphacephei.com/vosk/models and unpack as {} in the current folder.".format(model_path))
     exit (1)
 
 if not os.path.exists(spk_model_path):
-    print ("Please download the speaker model from https://github.com/alphacep/kaldi-android-demo/releases and unpack as {} in the current folder.".format(spk_model_path))
+    print ("Please download the speaker model from https://alphacephei.com/vosk/models and unpack as {} in the current folder.".format(spk_model_path))
     exit (1)
 
 wf = wave.open(sys.argv[1], "rb")
@@ -30,7 +30,7 @@ rec = KaldiRecognizer(model, spk_model, wf.getframerate())
 
 # We compare speakers with cosine distance. We can keep one or several fingerprints for the speaker in a database
 # to distingusih among users.
-spk_sig = [5.64308, 4.23898, 1.119433, -0.810904, 2.115443, 2.328436, 6.135152, 1.348195, 2.60771, 1.020717, 4.324225, -0.873012, 6.123375, 4.903791, 0.064803, 4.66212, 3.502724, 2.535861, 5.452417, 7.081769, -0.823969, -5.167974, 8.568919, 4.159035, 5.314441, 3.688272, 5.730379, 4.463213, 7.227232, 3.538961, 3.316218, 1.269628, -1.902378, 3.512679, -1.947611, -1.520158, 3.80928, -2.721601, 5.359588, 2.942463, -7.474174, 3.788054, 0.303426, 4.951366, 1.72281, -1.867125, -3.574615, 3.622509, 4.803109, 2.829714, 1.528521, 6.408293, 0.820131, 5.066522, 2.836125, 2.867029, 3.725267, 0.505927, 1.462984, 5.001863, -3.838309, -2.45902, 3.992581, 4.451616, 2.865211, -1.148313, 4.996399, -3.473454, 2.876967, 3.940124, 7.553079, 0.373356, 1.396561, 2.686691, 2.094895, 0.913796, -0.286909, 3.540179, 4.904687, 0.84554, 7.585956, 1.017081, 0.168355, 6.672327, 4.092033, -4.240158, -2.017081, -0.813043, 6.468298, 4.115041, 2.231936, 2.370055, 4.972295, 5.58382, 6.022872, 2.706988, 5.248096, -1.918003, 8.259204, -0.900911, 1.961962, 2.349709, 3.290093, 3.344172, 3.307027, 4.203372, -0.315103, 5.61919, -3.229496, 3.777309, 4.328595, 1.461014, 2.622894, 0.315525, 5.447259, 5.407609, 5.339016, 1.604555, 5.359932, 0.090242, 0.535306, 4.724705, 4.692502, 0.5783, -5.436688, -4.915511, 1.959807, 2.825248]
+spk_sig = [-1.110417,0.09703002,1.35658,0.7798632,-0.305457,-0.339204,0.6186931,-0.4521213,0.3982236,-0.004530723,0.7651616,0.6500852,-0.6664245,0.1361499,0.1358056,-0.2887807,-0.1280468,-0.8208137,-1.620276,-0.4628615,0.7870904,-0.105754,0.9739769,-0.3258137,-0.7322628,-0.6212429,-0.5531687,-0.7796484,0.7035915,1.056094,-0.4941756,-0.6521456,-0.2238328,-0.003737517,0.2165709,1.200186,-0.7737719,0.492015,1.16058,0.6135428,-0.7183084,0.3153541,0.3458071,-1.418189,-0.9624157,0.4168292,-1.627305,0.2742135,-0.6166027,0.1962581,-0.6406527,0.4372789,-0.4296024,0.4898657,-0.9531326,-0.2945702,0.7879696,-1.517101,-0.9344181,-0.5049928,-0.005040941,-0.4637912,0.8223695,-1.079849,0.8871287,-0.9732434,-0.5548235,1.879138,-1.452064,-0.1975368,1.55047,0.5941782,-0.52897,1.368219,0.6782904,1.202505,-0.9256122,-0.9718158,-0.9570228,-0.5563112,-1.19049,-1.167985,2.606804,-2.261825,0.01340385,0.2526799,-1.125458,-1.575991,-0.363153,0.3270262,1.485984,-1.769565,1.541829,0.7293826,0.1743717,-0.4759418,1.523451,-2.487134,-1.824067,-0.626367,0.7448186,-1.425648,0.3524166,-0.9903384,3.339342,0.4563958,-0.2876643,1.521635,0.9508078,-0.1398541,0.3867955,-0.7550205,0.6568405,0.09419366,-1.583935,1.306094,-0.3501927,0.1794427,-0.3768163,0.9683866,-0.2442541,-1.696921,-1.8056,-0.6803037,-1.842043,0.3069353,0.9070363,-0.486526]
 
 def cosine_dist(x, y):
     nx = np.array(x)
@@ -38,15 +38,19 @@ def cosine_dist(x, y):
     return 1 - np.dot(nx, ny) / np.linalg.norm(nx) / np.linalg.norm(ny)
 
 while True:
-    data = wf.readframes(1000)
+    data = wf.readframes(4000)
     if len(data) == 0:
         break
     if rec.AcceptWaveform(data):
         res = json.loads(rec.Result())
         print ("Text:", res['text'])
-        print ("Speaker distance:", cosine_dist(spk_sig, res['spk']))
+        print ("X-vector:", res['spk'])
+        print ("Speaker distance:", cosine_dist(spk_sig, res['spk']), "based on", res['spk_frames'], "frames")
+
+print ("Note that second distance is not very reliable because utterance is too short. Utterances longer than 4 seconds give better xvector")
 
 res = json.loads(rec.FinalResult())
 print ("Text:", res['text'])
-print ("Speaker distance:", cosine_dist(spk_sig, res['spk']))
+print ("X-vector:", res['spk'])
+print ("Speaker distance:", cosine_dist(spk_sig, res['spk']), "based on", res['spk_frames'], "frames")
 

python/example/test_srt.py

@@ -0,0 +1,49 @@
+#!/usr/bin/env python3
+
+from vosk import Model, KaldiRecognizer, SetLogLevel
+import sys
+import os
+import wave
+import subprocess
+import srt
+import json
+import datetime
+
+SetLogLevel(-1)
+
+if not os.path.exists("model"):
+    print ("Please download the model from https://alphacephei.com/vosk/models and unpack as 'model' in the current folder.")
+    exit (1)
+
+sample_rate=16000
+model = Model("model")
+rec = KaldiRecognizer(model, sample_rate)
+
+process = subprocess.Popen(['ffmpeg', '-loglevel', 'quiet', '-i',
+                            sys.argv[1],
+                            '-ar', str(sample_rate) , '-ac', '1', '-f', 's16le', '-'],
+                            stdout=subprocess.PIPE)
+
+
+def transcribe():
+    results = []
+    subs = []
+    while True:
+       data = process.stdout.read(4000)
+       if len(data) == 0:
+           break
+       if rec.AcceptWaveform(data):
+           results.append(rec.Result())
+    results.append(rec.FinalResult())
+
+    for i, res in enumerate(results):
+       jres = json.loads(res)
+       s = srt.Subtitle(index=i, 
+                   content=jres['text'], 
+                   start=datetime.timedelta(seconds=jres['result'][0]['start']), 
+                   end=datetime.timedelta(seconds=jres['result'][-1]['end']))
+       subs.append(s)
+    return subs
+
+print (srt.compose(transcribe()))
+

python/example/test_text.py

@@ -1,16 +1,16 @@
-#!/usr/bin/python3
+#!/usr/bin/env python3
 
 from vosk import Model, KaldiRecognizer
 import sys
 import json
 import os
 
-if not os.path.exists("model-en"):
-    print ("Please download the model from https://github.com/alphacep/kaldi-android-demo/releases and unpack as 'model-en' in the current folder.")
+if not os.path.exists("model"):
+    print ("Please download the model from https://alphacephei.com/vosk/models and unpack as 'model' in the current folder.")
     exit (1)
 
 
-model = Model("model-en")
+model = Model("model")
 
 # Large vocabulary free form recognition
 rec = KaldiRecognizer(model, 16000)
@@ -22,7 +22,7 @@ wf = open(sys.argv[1], "rb")
 wf.read(44) # skip header
 
 while True:
-    data = wf.read(2000)
+    data = wf.read(4000)
     if len(data) == 0:
         break
     if rec.AcceptWaveform(data):

python/example/test_words.py

@@ -1,12 +1,12 @@
-#!/usr/bin/python3
+#!/usr/bin/env python3
 
 from vosk import Model, KaldiRecognizer
 import sys
 import os
 import wave
 
-if not os.path.exists("model-en"):
-    print ("Please download the model from https://github.com/alphacep/kaldi-android-demo/releases and unpack as 'model-en' in the current folder.")
+if not os.path.exists("model"):
+    print ("Please download the model from https://alphacephei.com/vosk/models and unpack as 'model' in the current folder.")
     exit (1)
 
 wf = wave.open(sys.argv[1], "rb")
@@ -14,12 +14,13 @@ if wf.getnchannels() != 1 or wf.getsampwidth() != 2 or wf.getcomptype() != "NONE
     print ("Audio file must be WAV format mono PCM.")
     exit (1)
 
-model = Model("model-en")
-# You can also specify the possible word list
-rec = KaldiRecognizer(model, wf.getframerate(), "zero oh one two three four five six seven eight nine")
+model = Model("model")
+
+# You can also specify the possible word or phrase list as JSON list, the order doesn't have to be strict
+rec = KaldiRecognizer(model, wf.getframerate(), '["oh one two three four five six seven eight nine zero", "[unk]"]')
 
 while True:
-    data = wf.readframes(1000)
+    data = wf.readframes(4000)
     if len(data) == 0:
         break
     if rec.AcceptWaveform(data):

python/setup.py

@@ -1,22 +1,88 @@
 import os
+import sys
 import setuptools
-from cmake import *
+from setuptools import Extension
+from setuptools.command.build_py import build_py as _build_py
+import distutils.dir_util
+import distutils.log
+
+class build_py(_build_py):
+    def run(self):
+        self.run_command("build_ext")
+        return super().run()
+
+kaldi_root = os.getenv('KALDI_ROOT')
+kaldi_mkl = os.getenv('KALDI_MKL')
+source_path = os.getenv("VOSK_SOURCE", os.path.abspath(os.path.join(os.path.abspath(os.path.dirname(__file__)), "../src")))
+
+if kaldi_root == None:
+    print("Define KALDI_ROOT")
+    exit(1)
+
+distutils.log.set_verbosity(distutils.log.DEBUG)
+distutils.dir_util.copy_tree(
+    source_path,
+    "vosk",
+    update=1,
+    verbose=1)
 
 with open("README.md", "r") as fh:
     long_description = fh.read()
 
+kaldi_static_libs = ['src/online2/kaldi-online2.a',
+             'src/decoder/kaldi-decoder.a',
+             'src/ivector/kaldi-ivector.a',
+             'src/gmm/kaldi-gmm.a',
+             'src/nnet3/kaldi-nnet3.a',
+             'src/tree/kaldi-tree.a',
+             'src/feat/kaldi-feat.a',
+             'src/lat/kaldi-lat.a',
+             'src/lm/kaldi-lm.a',
+             'src/hmm/kaldi-hmm.a',
+             'src/transform/kaldi-transform.a',
+             'src/cudamatrix/kaldi-cudamatrix.a',
+             'src/matrix/kaldi-matrix.a',
+             'src/fstext/kaldi-fstext.a',
+             'src/util/kaldi-util.a',
+             'src/base/kaldi-base.a',
+             'tools/openfst/lib/libfst.a',
+             'tools/openfst/lib/libfstngram.a']
+kaldi_link_args = ['-s']
+kaldi_libraries = []
+
+if sys.platform.startswith('darwin'):
+    kaldi_link_args.extend(['-Wl,-undefined,dynamic_lookup', '-framework', 'Accelerate'])
+elif kaldi_mkl == "1":
+    kaldi_link_args.extend(['-L/opt/intel/mkl/lib/intel64', '-Wl,-rpath=/opt/intel/mkl/lib/intel64'])
+    kaldi_libraries.extend(['mkl_rt', 'mkl_intel_lp64', 'mkl_core', 'mkl_sequential'])
+else:
+    kaldi_static_libs.append('tools/OpenBLAS/libopenblas.a')
+    kaldi_libraries.append('gfortran')
+
+sources = ['kaldi_recognizer.cc', 'model.cc', 'spk_model.cc', 'vosk_api.cc', 'language_model.cc', 'vosk.i']
+
+vosk_ext = Extension('vosk._vosk',
+                    define_macros = [('FST_NO_DYNAMIC_LINKING', '1')],
+                    include_dirs = [kaldi_root + '/src', kaldi_root + '/tools/openfst/include', 'vosk'],
+                    swig_opts=['-outdir', 'vosk', '-c++'],
+                    libraries = kaldi_libraries,
+                    extra_objects = [kaldi_root + '/' + x for x in kaldi_static_libs],
+                    sources = ['vosk/' + x for x in sources],
+                    extra_link_args = kaldi_link_args,
+                    extra_compile_args = ['-std=c++11', '-Wno-sign-compare', '-Wno-unused-variable', '-Wno-unused-local-typedefs'])
+
 setuptools.setup(
-    name="vosk", # Replace with your own username
-    version="0.3.3",
+    name="vosk",
+    version="0.3.15",
     author="Alpha Cephei Inc",
     author_email="contact@alphacephei.com",
-    description="API for Kaldi and Vosk",
+    description="Offline open source speech recognition API based on Kaldi and Vosk",
     long_description=long_description,
     long_description_content_type="text/markdown",
     url="https://github.com/alphacep/vosk-api",
     packages=setuptools.find_packages(),
-    ext_modules=[CMakeExtension('_vosk', 'vosk')],
-    cmdclass={'build_ext': CMakeBuildExt, 'build_py' : CMakeBuildExtFirst},
+    ext_modules=[vosk_ext],
+    cmdclass = {'build_py' : build_py},
     classifiers=[
         'Programming Language :: Python :: 3',
         'License :: OSI Approved :: Apache Software License',
@@ -25,5 +91,5 @@ setuptools.setup(
         'Operating System :: MacOS :: MacOS X',
         'Topic :: Software Development :: Libraries :: Python Modules'
     ],
-    python_requires='>=3.4',
+    python_requires='>=3.5',
 )

python/test/transcribe_scp.py

@@ -0,0 +1,35 @@
+#!/usr/bin/env python3
+
+from multiprocessing.dummy import Pool
+from vosk import Model, KaldiRecognizer
+
+import sys
+import os
+import wave
+import json
+
+model = Model("model")
+
+def recognize(line):
+    uid, fn = line.split()
+    wf = wave.open(fn, "rb")
+    rec = KaldiRecognizer(model, wf.getframerate())
+
+    text = ""
+    while True:
+        data = wf.readframes(1000)
+        if len(data) == 0:
+            break
+        if rec.AcceptWaveform(data):
+            jres = json.loads(rec.Result())
+            text = text + " " + jres['text']
+    jres = json.loads(rec.FinalResult())
+    text = text + " " + jres['text']
+    return (uid + text)
+
+def main():
+    p = Pool(8)
+    texts = p.map(recognize, open(sys.argv[1]).readlines())
+    print ("\n".join(texts))
+
+main()

python/vosk/__init__.py

@@ -1 +1,2 @@
-from .vosk import KaldiRecognizer, Model, SpkModel
+from .vosk import KaldiRecognizer, Model, SpkModel, SetLogLevel
+

src/kaldi_recognizer.cc

@@ -1,4 +1,4 @@
-// Copyright 2019 Alpha Cephei Inc.
+// Copyright 2019-2020 Alpha Cephei Inc.
 //
 // Licensed under the Apache License, Version 2.0 (the "License");
 // you may not use this file except in compliance with the License.
@@ -16,124 +16,214 @@
 #include "json.h"
 #include "fstext/fstext-utils.h"
 #include "lat/sausages.h"
+#include "language_model.h"
 
 using namespace fst;
 using namespace kaldi::nnet3;
 
-KaldiRecognizer::KaldiRecognizer(Model &model, float sample_frequency) : model_(model), spk_model_(0), sample_frequency_(sample_frequency) {
+KaldiRecognizer::KaldiRecognizer(Model *model, float sample_frequency) : model_(model), spk_model_(0), sample_frequency_(sample_frequency) {
 
-    feature_pipeline_ = new kaldi::OnlineNnet2FeaturePipeline (model_.feature_info_);
-    silence_weighting_ = new kaldi::OnlineSilenceWeighting(*model_.trans_model_, model_.feature_info_.silence_weighting_config, 3);
+    model_->Ref();
 
+    feature_pipeline_ = new kaldi::OnlineNnet2FeaturePipeline (model_->feature_info_);
+    silence_weighting_ = new kaldi::OnlineSilenceWeighting(*model_->trans_model_, model_->feature_info_.silence_weighting_config, 3);
+
+    g_fst_ = NULL;
     decode_fst_ = NULL;
 
-    if (!model_.hclg_fst_) {
-        if (model_.hcl_fst_ && model_.g_fst_) {
-            decode_fst_ = LookaheadComposeFst(*model_.hcl_fst_, *model_.g_fst_, model_.disambig_);
+    if (!model_->hclg_fst_) {
+        if (model_->hcl_fst_ && model_->g_fst_) {
+            decode_fst_ = LookaheadComposeFst(*model_->hcl_fst_, *model_->g_fst_, model_->disambig_);
         } else {
             KALDI_ERR << "Can't create decoding graph";
         }
     }
 
-    decoder_ = new kaldi::SingleUtteranceNnet3Decoder(model_.nnet3_decoding_config_,
-            *model_.trans_model_,
-            *model_.decodable_info_,
-            model_.hclg_fst_ ? *model.hclg_fst_ : *decode_fst_,
+    decoder_ = new kaldi::SingleUtteranceNnet3Decoder(model_->nnet3_decoding_config_,
+            *model_->trans_model_,
+            *model_->decodable_info_,
+            model_->hclg_fst_ ? *model_->hclg_fst_ : *decode_fst_,
             feature_pipeline_);
 
-    frame_offset_ = 0;
-    input_finalized_ = false;
     spk_feature_ = NULL;
+
+    InitState();
+    InitRescoring();
 }
 
-KaldiRecognizer::KaldiRecognizer(Model &model, float sample_frequency, char const *grammar) : model_(model), spk_model_(0), sample_frequency_(sample_frequency)
+KaldiRecognizer::KaldiRecognizer(Model *model, float sample_frequency, char const *grammar) : model_(model), spk_model_(0), sample_frequency_(sample_frequency)
 {
-    feature_pipeline_ = new kaldi::OnlineNnet2FeaturePipeline (model_.feature_info_);
-    silence_weighting_ = new kaldi::OnlineSilenceWeighting(*model_.trans_model_, model_.feature_info_.silence_weighting_config, 3);
+    model_->Ref();
 
-    if (model_.hcl_fst_) {
-        g_fst_.AddState();
-        g_fst_.SetStart(0);
-        g_fst_.AddState();
-        g_fst_.SetFinal(1, fst::TropicalWeight::One());
-        g_fst_.AddArc(1, StdArc(0, 0, fst::TropicalWeight::One(), 0));
+    feature_pipeline_ = new kaldi::OnlineNnet2FeaturePipeline (model_->feature_info_);
+    silence_weighting_ = new kaldi::OnlineSilenceWeighting(*model_->trans_model_, model_->feature_info_.silence_weighting_config, 3);
 
-        // Create simple word loop FST
-        std::stringstream ss(grammar);
-        std::string token;
+    if (model_->hcl_fst_) {
+        json::JSON obj;
+        obj = json::JSON::Load(grammar);
 
-        while (std::getline(ss, token, ' ')) {
-            int32 id = model_.word_syms_->Find(token);
-            g_fst_.AddArc(0, StdArc(id, id, fst::TropicalWeight::One(), 1));
+        if (obj.length() <= 0) {
+            KALDI_WARN << "Expecting array of strings, got: '" << grammar << "'";
+        } else {
+            KALDI_LOG << obj;
+
+            LanguageModelOptions opts;
+
+            opts.ngram_order = 2;
+            opts.discount = 0.5;
+
+            LanguageModelEstimator estimator(opts);
+            for (int i = 0; i < obj.length(); i++) {
+                bool ok;
+                string line = obj[i].ToString(ok);
+                if (!ok) {
+                    KALDI_ERR << "Expecting array of strings, got: '" << obj << "'";
+                }
+
+                std::vector<int32> sentence;
+                stringstream ss(line);
+                string token;
+                while (getline(ss, token, ' ')) {
+                    int32 id = model_->word_syms_->Find(token);
+                    if (id == kNoSymbol) {
+                        KALDI_WARN << "Ignoring word missing in vocabulary: '" << token << "'";
+                    } else {
+                        sentence.push_back(id);
+                    }
+                }
+                estimator.AddCounts(sentence);
+            }
+            g_fst_ = new StdVectorFst();
+            estimator.Estimate(g_fst_);
+
+            decode_fst_ = LookaheadComposeFst(*model_->hcl_fst_, *g_fst_, model_->disambig_);
         }
-        ArcSort(&g_fst_, ILabelCompare<StdArc>());
-
-        decode_fst_ = LookaheadComposeFst(*model_.hcl_fst_, g_fst_, model_.disambig_);
     } else {
         decode_fst_ = NULL;
-        KALDI_ERR << "Can't create decoding graph";
+        KALDI_WARN << "Runtime graphs are not supported by this model";
     }
 
-    decoder_ = new kaldi::SingleUtteranceNnet3Decoder(model_.nnet3_decoding_config_,
-            *model_.trans_model_,
-            *model_.decodable_info_,
-            model_.hclg_fst_ ? *model.hclg_fst_ : *decode_fst_,
+    decoder_ = new kaldi::SingleUtteranceNnet3Decoder(model_->nnet3_decoding_config_,
+            *model_->trans_model_,
+            *model_->decodable_info_,
+            model_->hclg_fst_ ? *model_->hclg_fst_ : *decode_fst_,
             feature_pipeline_);
 
-    frame_offset_ = 0;
-    input_finalized_ = false;
     spk_feature_ = NULL;
+
+    InitState();
+    InitRescoring();
 }
 
+KaldiRecognizer::KaldiRecognizer(Model *model, SpkModel *spk_model, float sample_frequency) : model_(model), spk_model_(spk_model), sample_frequency_(sample_frequency) {
 
-KaldiRecognizer::KaldiRecognizer(Model &model, SpkModel *spk_model, float sample_frequency) : model_(model), spk_model_(spk_model), sample_frequency_(sample_frequency) {
-    feature_pipeline_ = new kaldi::OnlineNnet2FeaturePipeline (model_.feature_info_);
-    silence_weighting_ = new kaldi::OnlineSilenceWeighting(*model_.trans_model_, model_.feature_info_.silence_weighting_config, 3);
+    model_->Ref();
+    spk_model->Ref();
+
+    feature_pipeline_ = new kaldi::OnlineNnet2FeaturePipeline (model_->feature_info_);
+    silence_weighting_ = new kaldi::OnlineSilenceWeighting(*model_->trans_model_, model_->feature_info_.silence_weighting_config, 3);
 
     decode_fst_ = NULL;
+    g_fst_ = NULL;
 
-    if (!model_.hclg_fst_) {
-        if (model_.hcl_fst_ && model_.g_fst_) {
-            decode_fst_ = LookaheadComposeFst(*model_.hcl_fst_, *model_.g_fst_, model_.disambig_);
+    if (!model_->hclg_fst_) {
+        if (model_->hcl_fst_ && model_->g_fst_) {
+            decode_fst_ = LookaheadComposeFst(*model_->hcl_fst_, *model_->g_fst_, model_->disambig_);
         } else {
             KALDI_ERR << "Can't create decoding graph";
         }
     }
 
-    decoder_ = new kaldi::SingleUtteranceNnet3Decoder(model_.nnet3_decoding_config_,
-            *model_.trans_model_,
-            *model_.decodable_info_,
-            model_.hclg_fst_ ? *model.hclg_fst_ : *decode_fst_,
+    decoder_ = new kaldi::SingleUtteranceNnet3Decoder(model_->nnet3_decoding_config_,
+            *model_->trans_model_,
+            *model_->decodable_info_,
+            model_->hclg_fst_ ? *model_->hclg_fst_ : *decode_fst_,
             feature_pipeline_);
 
-    frame_offset_ = 0;
-    input_finalized_ = false;
-
     spk_feature_ = new OnlineMfcc(spk_model_->spkvector_mfcc_opts);
+
+    InitState();
+    InitRescoring();
 }
 
 KaldiRecognizer::~KaldiRecognizer() {
+    delete decoder_;
     delete feature_pipeline_;
     delete silence_weighting_;
-    delete decoder_;
+    delete g_fst_;
     delete decode_fst_;
     delete spk_feature_;
+    delete lm_fst_;
+
+    model_->Unref();
+    if (spk_model_)
+         spk_model_->Unref();
+}
+
+void KaldiRecognizer::InitState()
+{
+    frame_offset_ = 0;
+    samples_processed_ = 0;
+    samples_round_start_ = 0;
+
+    state_ = RECOGNIZER_INITIALIZED;
+}
+
+void KaldiRecognizer::InitRescoring()
+{
+    if (model_->std_lm_fst_) {
+        fst::CacheOptions cache_opts(true, 50000);
+        fst::MapFstOptions mapfst_opts(cache_opts);
+        fst::StdToLatticeMapper<kaldi::BaseFloat> mapper;
+        lm_fst_ = new fst::MapFst<fst::StdArc, kaldi::LatticeArc, fst::StdToLatticeMapper<kaldi::BaseFloat> >(*model_->std_lm_fst_, mapper, mapfst_opts);
+    } else {
+        lm_fst_ = NULL;
+    }
 }
 
 void KaldiRecognizer::CleanUp()
 {
     delete silence_weighting_;
-    silence_weighting_ = new kaldi::OnlineSilenceWeighting(*model_.trans_model_, model_.feature_info_.silence_weighting_config, 3);
+    silence_weighting_ = new kaldi::OnlineSilenceWeighting(*model_->trans_model_, model_->feature_info_.silence_weighting_config, 3);
 
-    frame_offset_ += decoder_->NumFramesDecoded();
-    decoder_->InitDecoding(frame_offset_);
+    if (decoder_)
+       frame_offset_ += decoder_->NumFramesDecoded();
+
+    // Each 10 minutes we drop the pipeline to save frontend memory in continuous processing
+    // here we drop few frames remaining in the feature pipeline but hope it will not
+    // cause a huge accuracy drop since it happens not very frequently.
+
+    // Also restart if we retrieved final result already
+
+    if (decoder_ == NULL || state_ == RECOGNIZER_FINALIZED || frame_offset_ > 20000) {
+        samples_round_start_ += samples_processed_;
+        samples_processed_ = 0;
+        frame_offset_ = 0;
+
+        delete decoder_;
+        delete feature_pipeline_;
+
+        feature_pipeline_ = new kaldi::OnlineNnet2FeaturePipeline (model_->feature_info_);
+        decoder_ = new kaldi::SingleUtteranceNnet3Decoder(model_->nnet3_decoding_config_,
+            *model_->trans_model_,
+            *model_->decodable_info_,
+            model_->hclg_fst_ ? *model_->hclg_fst_ : *decode_fst_,
+            feature_pipeline_);
+
+        if (spk_model_) {
+            delete spk_feature_;
+            spk_feature_ = new OnlineMfcc(spk_model_->spkvector_mfcc_opts);
+        }
+    } else {
+        decoder_->InitDecoding(frame_offset_);
+    }
 }
 
 void KaldiRecognizer::UpdateSilenceWeights()
 {
     if (silence_weighting_->Active() && feature_pipeline_->NumFramesReady() > 0 &&
         feature_pipeline_->IvectorFeature() != NULL) {
-        std::vector<std::pair<int32, BaseFloat> > delta_weights;
+        vector<pair<int32, BaseFloat> > delta_weights;
         silence_weighting_->ComputeCurrentTraceback(decoder_->Decoder());
         silence_weighting_->GetDeltaWeights(feature_pipeline_->NumFramesReady(),
                                           frame_offset_ * 3,
@@ -171,27 +261,32 @@ bool KaldiRecognizer::AcceptWaveform(const float *fdata, int len)
 
 bool KaldiRecognizer::AcceptWaveform(Vector<BaseFloat> &wdata)
 {
-    if (input_finalized_) {
+    // Cleanup if we finalized previous utterance or the whole feature pipeline
+    if (!(state_ == RECOGNIZER_RUNNING || state_ == RECOGNIZER_INITIALIZED)) {
         CleanUp();
-        input_finalized_ = false;
     }
+    state_ = RECOGNIZER_RUNNING;
 
-    feature_pipeline_->AcceptWaveform(sample_frequency_, wdata);
-    UpdateSilenceWeights();
-    decoder_->AdvanceDecoding();
+    int step = static_cast<int>(sample_frequency_ * 0.2);
+    for (int i = 0; i < wdata.Dim(); i+= step) {
+        SubVector<BaseFloat> r = wdata.Range(i, std::min(step, wdata.Dim() - i));
+        feature_pipeline_->AcceptWaveform(sample_frequency_, r);
+        UpdateSilenceWeights();
+        decoder_->AdvanceDecoding();
+    }
+    samples_processed_ += wdata.Dim();
 
     if (spk_feature_) {
         spk_feature_->AcceptWaveform(sample_frequency_, wdata);
     }
 
-    if (decoder_->EndpointDetected(model_.endpoint_config_)) {
+    if (decoder_->EndpointDetected(model_->endpoint_config_)) {
         return true;
     }
 
     return false;
 }
 
-
 // Computes an xvector from a chunk of speech features.
 static void RunNnetComputation(const MatrixBase<BaseFloat> &features,
     const nnet3::Nnet &nnet, nnet3::CachingOptimizingCompiler *compiler,
@@ -208,7 +303,7 @@ static void RunNnetComputation(const MatrixBase<BaseFloat> &features,
     output_spec.indexes.resize(1);
     request.outputs.resize(1);
     request.outputs[0].Swap(&output_spec);
-    std::shared_ptr<const nnet3::NnetComputation> computation = compiler->Compile(request);
+    shared_ptr<const nnet3::NnetComputation> computation = compiler->Compile(request);
     nnet3::Nnet *nnet_to_update = NULL;  // we're not doing any update.
     nnet3::NnetComputer computer(nnet3::NnetComputeOptions(), *computation,
                     nnet, nnet_to_update);
@@ -221,17 +316,48 @@ static void RunNnetComputation(const MatrixBase<BaseFloat> &features,
     xvector->CopyFromVec(cu_output.Row(0));
 }
 
+#define MIN_SPK_FEATS 50
 
-void KaldiRecognizer::GetSpkVector(Vector<BaseFloat> &xvector)
+bool KaldiRecognizer::GetSpkVector(Vector<BaseFloat> &out_xvector, int *num_spk_frames)
 {
+    vector<int32> nonsilence_frames;
+    if (silence_weighting_->Active() && feature_pipeline_->NumFramesReady() > 0) {
+        silence_weighting_->ComputeCurrentTraceback(decoder_->Decoder(), true);
+        silence_weighting_->GetNonsilenceFrames(feature_pipeline_->NumFramesReady(),
+                                          frame_offset_ * 3,
+                                          &nonsilence_frames);
+    }
+
     int num_frames = spk_feature_->NumFramesReady() - frame_offset_ * 3;
     Matrix<BaseFloat> mfcc(num_frames, spk_feature_->Dim());
+
+    // Not very efficient, would be nice to have faster search
+    int num_nonsilence_frames = 0;
+    Vector<BaseFloat> feat(spk_feature_->Dim());
+
     for (int i = 0; i < num_frames; ++i) {
-       Vector<BaseFloat> feat(spk_feature_->Dim());
+       if (std::find(nonsilence_frames.begin(),
+                     nonsilence_frames.end(), i / 3) == nonsilence_frames.end()) {
+           continue;
+       }
+
        spk_feature_->GetFrame(i + frame_offset_ * 3, &feat);
-       mfcc.CopyRowFromVec(feat, i);
+       mfcc.CopyRowFromVec(feat, num_nonsilence_frames);
+       num_nonsilence_frames++;
     }
+
+    *num_spk_frames = num_nonsilence_frames;
+
+    // Don't extract vector if not enough data
+    if (num_nonsilence_frames < MIN_SPK_FEATS) {
+        return false;
+    }
+
+    mfcc.Resize(num_nonsilence_frames, spk_feature_->Dim(), kCopyData);
+
     SlidingWindowCmnOptions cmvn_opts;
+    cmvn_opts.center = true;
+    cmvn_opts.cmn_window = 300;
     Matrix<BaseFloat> features(mfcc.NumRows(), mfcc.NumCols(), kUndefined);
     SlidingWindowCmn(cmvn_opts, mfcc, &features);
 
@@ -239,105 +365,178 @@ void KaldiRecognizer::GetSpkVector(Vector<BaseFloat> &xvector)
     nnet3::CachingOptimizingCompilerOptions compiler_config;
     nnet3::CachingOptimizingCompiler compiler(spk_model_->speaker_nnet, opts.optimize_config, compiler_config);
 
+    Vector<BaseFloat> xvector;
     RunNnetComputation(features, spk_model_->speaker_nnet, &compiler, &xvector);
+
+    // Whiten the vector with global mean and transform and normalize mean
+    xvector.AddVec(-1.0, spk_model_->mean);
+
+    out_xvector.Resize(spk_model_->transform.NumRows(), kSetZero);
+    out_xvector.AddMatVec(1.0, spk_model_->transform, kNoTrans, xvector, 0.0);
+
+    BaseFloat norm = out_xvector.Norm(2.0);
+    BaseFloat ratio = norm / sqrt(out_xvector.Dim()); // how much larger it is
+                                                  // than it would be, in
+                                                  // expectation, if normally
+    out_xvector.Scale(1.0 / ratio);
+
+    return true;
 }
 
-
-std::string KaldiRecognizer::Result()
+const char* KaldiRecognizer::GetResult()
 {
-
-    if (!input_finalized_) {
-        decoder_->FinalizeDecoding();
-        input_finalized_ = true;
-    }
-
     if (decoder_->NumFramesDecoded() == 0) {
-        return "{\"text\": \"\"}";
+        return StoreReturn("{\"text\": \"\"}");
     }
 
     kaldi::CompactLattice clat;
     decoder_->GetLattice(true, &clat);
-    fst::ScaleLattice(fst::LatticeScale(8.0, 10.0), &clat);
 
+    if (model_->std_lm_fst_) {
+        Lattice lat1;
+
+        ConvertLattice(clat, &lat1);
+        fst::ScaleLattice(fst::GraphLatticeScale(-1.0), &lat1);
+        fst::ArcSort(&lat1, fst::OLabelCompare<kaldi::LatticeArc>());
+        kaldi::Lattice composed_lat;
+        fst::Compose(lat1, *lm_fst_, &composed_lat);
+        fst::Invert(&composed_lat);
+        kaldi::CompactLattice determinized_lat;
+        DeterminizeLattice(composed_lat, &determinized_lat);
+        fst::ScaleLattice(fst::GraphLatticeScale(-1), &determinized_lat);
+        fst::ArcSort(&determinized_lat, fst::OLabelCompare<kaldi::CompactLatticeArc>());
+
+        kaldi::ConstArpaLmDeterministicFst const_arpa_fst(model_->const_arpa_);
+        kaldi::CompactLattice composed_clat;
+        kaldi::ComposeCompactLatticeDeterministic(determinized_lat, &const_arpa_fst, &composed_clat);
+        kaldi::Lattice composed_lat1;
+        ConvertLattice(composed_clat, &composed_lat1);
+        fst::Invert(&composed_lat1);
+        DeterminizeLattice(composed_lat1, &clat);
+    }
+
+    fst::ScaleLattice(fst::GraphLatticeScale(0.9), &clat); // Apply rescoring weight
     CompactLattice aligned_lat;
-    if (model_.winfo_) {
-        WordAlignLattice(clat, *model_.trans_model_, *model_.winfo_, 0, &aligned_lat);
+    if (model_->winfo_) {
+        WordAlignLattice(clat, *model_->trans_model_, *model_->winfo_, 0, &aligned_lat);
     } else {
         aligned_lat = clat;
     }
 
     MinimumBayesRisk mbr(aligned_lat);
-    const std::vector<BaseFloat> &conf = mbr.GetOneBestConfidences();
-    const std::vector<int32> &words = mbr.GetOneBest();
-    const std::vector<std::pair<BaseFloat, BaseFloat> > &times =
+    const vector<BaseFloat> &conf = mbr.GetOneBestConfidences();
+    const vector<int32> &words = mbr.GetOneBest();
+    const vector<pair<BaseFloat, BaseFloat> > &times =
           mbr.GetOneBestTimes();
 
     int size = words.size();
 
     json::JSON obj;
-    std::stringstream text;
+    stringstream text;
 
     // Create JSON object
     for (int i = 0; i < size; i++) {
         json::JSON word;
-        word["word"] = model_.word_syms_->Find(words[i]);
-        word["start"] = (frame_offset_ + times[i].first) * 0.03;
-        word["end"] = (frame_offset_ + times[i].second) * 0.03;
+        word["word"] = model_->word_syms_->Find(words[i]);
+        word["start"] = samples_round_start_ / sample_frequency_ + (frame_offset_ + times[i].first) * 0.03;
+        word["end"] = samples_round_start_ / sample_frequency_ + (frame_offset_ + times[i].second) * 0.03;
         word["conf"] = conf[i];
         obj["result"].append(word);
 
         if (i) {
             text << " ";
         }
-        text << model_.word_syms_->Find(words[i]);
+        text << model_->word_syms_->Find(words[i]);
     }
     obj["text"] = text.str();
 
     if (spk_model_) {
         Vector<BaseFloat> xvector;
-        GetSpkVector(xvector);
-        for (int i = 0; i < xvector.Dim(); i++) {
-            obj["spk"].append(xvector(i));
+        int num_spk_frames;
+        if (GetSpkVector(xvector, &num_spk_frames)) {
+            for (int i = 0; i < xvector.Dim(); i++) {
+                obj["spk"].append(xvector(i));
+            }
+            obj["spk_frames"] = num_spk_frames;
         }
     }
 
-    return obj.dump();
+    return StoreReturn(obj.dump());
 }
 
-std::string KaldiRecognizer::PartialResult()
+
+const char* KaldiRecognizer::PartialResult()
 {
+    if (state_ != RECOGNIZER_RUNNING) {
+        return StoreReturn("{\"text\": \"\"}");
+    }
+
     json::JSON res;
+
     if (decoder_->NumFramesDecoded() == 0) {
         res["partial"] = "";
-        return res.dump();
+        return StoreReturn(res.dump());
     }
 
     kaldi::Lattice lat;
     decoder_->GetBestPath(false, &lat);
-    std::vector<kaldi::int32> alignment, words;
+    vector<kaldi::int32> alignment, words;
     LatticeWeight weight;
     GetLinearSymbolSequence(lat, &alignment, &words, &weight);
 
-    std::ostringstream text;
+    ostringstream text;
     for (size_t i = 0; i < words.size(); i++) {
         if (i) {
             text << " ";
         }
-        text << model_.word_syms_->Find(words[i]);
+        text << model_->word_syms_->Find(words[i]);
     }
     res["partial"] = text.str();
 
-    return res.dump();
+    return StoreReturn(res.dump());
 }
 
-std::string KaldiRecognizer::FinalResult()
+const char* KaldiRecognizer::Result()
 {
-    if (!input_finalized_) {
-        feature_pipeline_->InputFinished();
-        UpdateSilenceWeights();
-        decoder_->AdvanceDecoding();
-        decoder_->FinalizeDecoding();
-        input_finalized_ = true;
+    if (state_ != RECOGNIZER_RUNNING) {
+        return StoreReturn("{\"text\": \"\"}");
     }
-    return Result();
+    decoder_->FinalizeDecoding();
+    state_ = RECOGNIZER_ENDPOINT;
+    return GetResult();
+}
+
+const char* KaldiRecognizer::FinalResult()
+{
+    if (state_ != RECOGNIZER_RUNNING) {
+        return StoreReturn("{\"text\": \"\"}");
+    }
+
+    feature_pipeline_->InputFinished();
+    UpdateSilenceWeights();
+    decoder_->AdvanceDecoding();
+    decoder_->FinalizeDecoding();
+    state_ = RECOGNIZER_FINALIZED;
+    GetResult();
+
+    // Free some memory while we are finalized, next
+    // iteration will reinitialize them anyway
+    delete decoder_;
+    delete feature_pipeline_;
+    delete silence_weighting_;
+    delete spk_feature_;
+
+    feature_pipeline_ = NULL;
+    silence_weighting_ = NULL;
+    decoder_ = NULL;
+    spk_feature_ = NULL;
+
+    return last_result_.c_str();
+}
+
+// Store result in recognizer and return as const string
+const char *KaldiRecognizer::StoreReturn(const string &res)
+{
+    last_result_ = res;
+    return last_result_.c_str();
 }

src/kaldi_recognizer.h

@@ -12,6 +12,9 @@
 // See the License for the specific language governing permissions and
 // limitations under the License.
 
+#ifndef VOSK_KALDI_RECOGNIZER_H
+#define VOSK_KALDI_RECOGNIZER_H
+
 #include "base/kaldi-common.h"
 #include "util/common-utils.h"
 #include "fstext/fstext-lib.h"
@@ -29,36 +32,56 @@
 
 using namespace kaldi;
 
+enum KaldiRecognizerState {
+    RECOGNIZER_INITIALIZED,
+    RECOGNIZER_RUNNING,
+    RECOGNIZER_ENDPOINT,
+    RECOGNIZER_FINALIZED
+};
+
 class KaldiRecognizer {
     public:
-        KaldiRecognizer(Model &model, float sample_frequency);
-        KaldiRecognizer(Model &model, SpkModel *spk_model, float sample_frequency);
-        KaldiRecognizer(Model &model, float sample_frequency, char const *grammar);
+        KaldiRecognizer(Model *model, float sample_frequency);
+        KaldiRecognizer(Model *model, SpkModel *spk_model, float sample_frequency);
+        KaldiRecognizer(Model *model, float sample_frequency, char const *grammar);
         ~KaldiRecognizer();
         bool AcceptWaveform(const char *data, int len);
         bool AcceptWaveform(const short *sdata, int len);
         bool AcceptWaveform(const float *fdata, int len);
-        std::string Result();
-        std::string FinalResult();
-        std::string PartialResult();
+        const char* Result();
+        const char* FinalResult();
+        const char* PartialResult();
 
     private:
+        void InitState();
+        void InitRescoring();
         void CleanUp();
         void UpdateSilenceWeights();
         bool AcceptWaveform(Vector<BaseFloat> &wdata);
-        void GetSpkVector(Vector<BaseFloat> &xvector);
+        bool GetSpkVector(Vector<BaseFloat> &out_xvector, int *frames);
+        const char *GetResult();
+        const char *StoreReturn(const string &res);
 
-        Model &model_;
+        Model *model_;
         SingleUtteranceNnet3Decoder *decoder_;
         fst::LookaheadFst<fst::StdArc, int32> *decode_fst_;
-        fst::StdVectorFst g_fst_; // dynamically constructed grammar
+        fst::StdVectorFst *g_fst_; // dynamically constructed grammar
         OnlineNnet2FeaturePipeline *feature_pipeline_;
         OnlineSilenceWeighting *silence_weighting_;
 
         SpkModel *spk_model_;
         OnlineBaseFeature *spk_feature_;
 
+        fst::MapFst<fst::StdArc, kaldi::LatticeArc, fst::StdToLatticeMapper<kaldi::BaseFloat> > *lm_fst_;
+
         float sample_frequency_;
         int32 frame_offset_;
-        bool input_finalized_;
+
+        int64 samples_processed_;
+        int64 samples_round_start_;
+
+        KaldiRecognizerState state_;
+        string last_result_;
 };
+
+#endif /* VOSK_KALDI_RECOGNIZER_H */

src/language_model.cc

@@ -0,0 +1,211 @@
+// Copyright      2015   Johns Hopkins University (author: Daniel Povey)
+
+// See ../../COPYING for clarification regarding multiple authors
+//
+// Licensed under the Apache License, Version 2.0 (the "License");
+// you may not use this file except in compliance with the License.
+// You may obtain a copy of the License at
+//
+//  http://www.apache.org/licenses/LICENSE-2.0
+//
+// THIS CODE IS PROVIDED *AS IS* BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
+// KIND, EITHER EXPRESS OR IMPLIED, INCLUDING WITHOUT LIMITATION ANY IMPLIED
+// WARRANTIES OR CONDITIONS OF TITLE, FITNESS FOR A PARTICULAR PURPOSE,
+// MERCHANTABLITY OR NON-INFRINGEMENT.
+// See the Apache 2 License for the specific language governing permissions and
+// limitations under the License.
+
+// A modified version from chain/language-model.cc for static backoff
+
+#include <algorithm>
+#include <numeric>
+
+#include "language_model.h"
+
+using namespace kaldi;
+
+void LanguageModelEstimator::AddCounts(const std::vector<int32> &sentence) {
+  KALDI_ASSERT(opts_.ngram_order >= 2 && "--ngram-order must be >= 2");
+  int32 order = opts_.ngram_order;
+  // 0 is used for left-context at the beginning of the file.. treat it as BOS.
+  std::vector<int32> history(0);
+  std::vector<int32>::const_iterator iter = sentence.begin(),
+      end = sentence.end();
+  for (; iter != end; ++iter) {
+    KALDI_ASSERT(*iter != 0);
+    IncrementCount(history, *iter);
+    history.push_back(*iter);
+    if (history.size() >= order)
+      history.erase(history.begin());
+  }
+  // Probability of end of sentence.  This will end up getting ignored later, but
+  // it still makes a difference for probability-normalization reasons.
+  IncrementCount(history, 0);
+}
+
+void LanguageModelEstimator::IncrementCount(const std::vector<int32> &history,
+                                            int32 next_phone) {
+  int32 lm_state_index = FindOrCreateLmStateIndexForHistory(history);
+  if (lm_states_[lm_state_index].tot_count == 0) {
+    num_active_lm_states_++;
+  }
+  lm_states_[lm_state_index].AddCount(next_phone, 1);
+}
+
+void LanguageModelEstimator::SetParentCounts() {
+  int32 num_lm_states = lm_states_.size();
+  for (int32 l = 0; l < num_lm_states; l++) {
+    int32 l_iter = lm_states_[l].backoff_lmstate_index;
+    while (l_iter != -1) {
+      lm_states_[l_iter].Add(lm_states_[l]);
+      l_iter = lm_states_[l_iter].backoff_lmstate_index;
+    }
+  }
+}
+
+int32 LanguageModelEstimator::FindLmStateIndexForHistory(
+    const std::vector<int32> &hist) const {
+  MapType::const_iterator iter = hist_to_lmstate_index_.find(hist);
+  if (iter == hist_to_lmstate_index_.end())
+    return -1;
+  else
+    return iter->second;
+}
+
+int32 LanguageModelEstimator::FindNonzeroLmStateIndexForHistory(
+    std::vector<int32> hist) const {
+  while (1) {
+    int32 l = FindLmStateIndexForHistory(hist);
+    if (l == -1 || lm_states_[l].tot_count == 0) {
+      // no such state or state has zero count.
+      if (hist.empty())
+        KALDI_ERR << "Error looking up LM state index for history "
+                  << "(likely code bug)";
+      hist.erase(hist.begin());  // back off.
+    } else {
+      return l;
+    }
+  }
+}
+
+int32 LanguageModelEstimator::FindOrCreateLmStateIndexForHistory(
+    const std::vector<int32> &hist) {
+  MapType::const_iterator iter = hist_to_lmstate_index_.find(hist);
+  if (iter != hist_to_lmstate_index_.end())
+    return iter->second;
+  int32 ans = lm_states_.size();  // index of next element
+  // next statement relies on default construct of LmState.
+  lm_states_.resize(lm_states_.size() + 1);
+  lm_states_.back().history = hist;
+  hist_to_lmstate_index_[hist] = ans;
+
+  // make sure backoff_lmstate_index is set
+  if (hist.size() > 0) {
+    std::vector<int32> backoff_hist(hist.begin() + 1,
+                                    hist.end());
+    int32 backoff_lm_state = FindOrCreateLmStateIndexForHistory(backoff_hist);
+    lm_states_[ans].backoff_lmstate_index = backoff_lm_state;
+  }
+  return ans;
+}
+
+void LanguageModelEstimator::LmState::AddCount(int32 phone, int32 count) {
+  std::map<int32, int32>::iterator iter = phone_to_count.find(phone);
+  if (iter == phone_to_count.end())
+    phone_to_count[phone] = count;
+  else
+    iter->second += count;
+  tot_count += count;
+}
+
+void LanguageModelEstimator::LmState::Add(const LmState &other) {
+  KALDI_ASSERT(&other != this);
+  std::map<int32, int32>::const_iterator iter = other.phone_to_count.begin(),
+      end = other.phone_to_count.end();
+  for (; iter != end; ++iter)
+    AddCount(iter->first, iter->second);
+}
+
+int32 LanguageModelEstimator::AssignFstStates() {
+  int32 num_lm_states = lm_states_.size();
+  int32 current_fst_state = 0;
+  for (int32 l = 0; l < num_lm_states; l++) {
+    if (lm_states_[l].tot_count != 0) {
+      lm_states_[l].fst_state = current_fst_state++;
+    }
+  }
+  KALDI_ASSERT(current_fst_state == num_active_lm_states_);
+  return current_fst_state;
+}
+
+void LanguageModelEstimator::Estimate(fst::StdVectorFst *fst) {
+  KALDI_LOG << "Estimating language model with ngram-order="
+            << opts_.ngram_order << ", discount="
+            << opts_.discount;
+  SetParentCounts();
+  int32 num_fst_states = AssignFstStates();
+  OutputToFst(num_fst_states, fst);
+}
+
+int32 LanguageModelEstimator::FindInitialFstState() const {
+  std::vector<int32> history(0);
+  int32 l = FindNonzeroLmStateIndexForHistory(history);
+  KALDI_ASSERT(l != -1 && lm_states_[l].fst_state != -1);
+  return lm_states_[l].fst_state;
+}
+
+void LanguageModelEstimator::OutputToFst(
+    int32 num_states,
+    fst::StdVectorFst *fst) const {
+  KALDI_ASSERT(num_states == num_active_lm_states_);
+  fst->DeleteStates();
+  for (int32 i = 0; i < num_states; i++)
+    fst->AddState();
+  fst->SetStart(FindInitialFstState());
+
+  int64 tot_count = 0;
+  double tot_logprob = 0.0;
+
+  int32 num_lm_states = lm_states_.size();
+  // note: not all lm-states end up being 'active'.
+  for (int32 l = 0; l < num_lm_states; l++) {
+    const LmState &lm_state = lm_states_[l];
+    if (lm_state.fst_state == -1) {
+      continue;
+    }
+    int32 state_count = lm_state.tot_count;
+    KALDI_ASSERT(state_count != 0);
+    std::map<int32, int32>::const_iterator
+        iter = lm_state.phone_to_count.begin(),
+        end = lm_state.phone_to_count.end();
+    for (; iter != end; ++iter) {
+      int32 phone = iter->first, count = iter->second;
+      BaseFloat logprob = log(count * opts_.discount / state_count);
+      tot_count += count;
+      tot_logprob += logprob * count;
+      if (phone == 0) {  // Go to final state
+        fst->SetFinal(lm_state.fst_state, fst::TropicalWeight(-logprob));
+      } else {  // It becomes a transition.
+        std::vector<int32> next_history(lm_state.history);
+        next_history.push_back(phone);
+        int32 dest_lm_state = FindNonzeroLmStateIndexForHistory(next_history),
+            dest_fst_state = lm_states_[dest_lm_state].fst_state;
+        KALDI_ASSERT(dest_fst_state != -1);
+        fst->AddArc(lm_state.fst_state,
+                    fst::StdArc(phone, phone, fst::TropicalWeight(-logprob),
+                                dest_fst_state));
+      }
+    }
+    if (lm_state.backoff_lmstate_index >= 0) {
+      fst->AddArc(lm_state.fst_state, fst::StdArc(0, 0, fst::TropicalWeight(-log(1 - opts_.discount)), lm_states_[lm_state.backoff_lmstate_index].fst_state));
+    }
+  }
+  fst::Connect(fst);
+  // Make sure that Connect does not delete any states.
+  int32 num_states_connected = fst->NumStates();
+  KALDI_ASSERT(num_states_connected == num_states);
+  // arc-sort.  ilabel or olabel doesn't matter, it's an acceptor.
+  fst::ArcSort(fst, fst::ILabelCompare<fst::StdArc>());
+  KALDI_LOG << "Created language model with " << num_states
+            << " states and " << fst::NumArcs(*fst) << " arcs.";
+}

src/language_model.h

@@ -0,0 +1,150 @@
+// Copyright      2015  Johns Hopkins University (Author: Daniel Povey)
+
+// See ../../COPYING for clarification regarding multiple authors
+//
+// Licensed under the Apache License, Version 2.0 (the "License");
+// you may not use this file except in compliance with the License.
+// You may obtain a copy of the License at
+//
+//  http://www.apache.org/licenses/LICENSE-2.0
+//
+// THIS CODE IS PROVIDED *AS IS* BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
+// KIND, EITHER EXPRESS OR IMPLIED, INCLUDING WITHOUT LIMITATION ANY IMPLIED
+// WARRANTIES OR CONDITIONS OF TITLE, FITNESS FOR A PARTICULAR PURPOSE,
+// MERCHANTABLITY OR NON-INFRINGEMENT.
+// See the Apache 2 License for the specific language governing permissions and
+// limitations under the License.
+
+
+#ifndef VOSK_LANGUAGE_MODEL_H
+#define VOSK_LANGUAGE_MODEL_H
+
+#include <vector>
+#include <map>
+
+#include "base/kaldi-common.h"
+#include "util/common-utils.h"
+#include "fstext/fstext-lib.h"
+#include "lat/kaldi-lattice.h"
+
+using namespace kaldi;
+
+// Very simply lm construction with absolute discounting
+
+struct LanguageModelOptions {
+  int32 ngram_order;  // you might want to tune this
+  BaseFloat discount; // discount for backoff
+
+  LanguageModelOptions():
+      ngram_order(3),
+      discount(0.5)
+      { }
+
+  void Register(OptionsItf *opts) {
+    opts->Register("ngram-order", &ngram_order, "n-gram order for the phone "
+                   "language model used for the 'denominator model'");
+    opts->Register("discount", &discount, "Discount for backoff");
+  }
+};
+
+class LanguageModelEstimator {
+ public:
+  LanguageModelEstimator(LanguageModelOptions &opts): opts_(opts),
+                                                      num_active_lm_states_(0) {
+    KALDI_ASSERT(opts.ngram_order >= 1);
+  }
+
+  // Adds counts for this sentence.  Basically does: for each n-gram in the
+  // sentence, count[n-gram] += 1.  The only constraint on 'sentence' is that it
+  // should contain no zeros.
+  void AddCounts(const std::vector<int32> &sentence);
+
+  // Estimates the LM and outputs it as an FST.  Note: there is
+  // no concept here of backoff arcs.
+  void Estimate(fst::StdVectorFst *fst);
+
+ protected:
+  struct LmState {
+    // the phone history associated with this state (length can vary).
+    std::vector<int32> history;
+
+    // maps from
+    std::map<int32, int32> phone_to_count;
+
+    // total count of this state.  As we back off states to lower-order states
+    // (and note that this is a hard backoff where we completely remove un-needed
+    // states) this tot_count may become zero.
+    int32 tot_count;
+
+    // LM-state index of the backoff LM state (if it exists, else -1)...
+    // provided for convenience.
+    int32 backoff_lmstate_index;
+
+    // this is only set after we decide on the FST state numbering (at the end).
+    // If not set, it's -1.
+    int32 fst_state;
+
+    void AddCount(int32 phone, int32 count);
+
+    // Add the contents of another LmState.
+    void Add(const LmState &other);
+
+    LmState(): tot_count(0), backoff_lmstate_index(-1),
+               fst_state(-1) { }
+    LmState(const LmState &other):
+        history(other.history), phone_to_count(other.phone_to_count),
+        tot_count(other.tot_count),
+        backoff_lmstate_index(other.backoff_lmstate_index),
+        fst_state(other.fst_state) { }
+  };
+
+  // maps from history to int32
+  typedef unordered_map<std::vector<int32>, int32, VectorHasher<int32> > MapType;
+
+  LanguageModelOptions opts_;
+
+  MapType hist_to_lmstate_index_;
+  std::vector<LmState> lm_states_;  // indexed by lmstate_index, the LmStates.
+
+  // Keeps track of the number of lm states that have nonzero counts.
+  int32 num_active_lm_states_;
+
+
+  // adds the counts for this ngram (called from AddCounts()).
+  inline void IncrementCount(const std::vector<int32> &history,
+                             int32 next_phone);
+
+  // sets up tot_count_with_parents in all the lm-states
+  void SetParentCounts();
+
+  // Finds and returns an LM-state index for a history -- or -1 if it doesn't
+  // exist.  No backoff is done.
+  int32 FindLmStateIndexForHistory(const std::vector<int32> &hist) const;
+
+  // Finds and returns an LM-state index for a history -- and creates one if
+  // it doesn't exist -- and also creates any backoff states needed, down
+  // to history-length no_prune_ngram_order - 1.
+  int32 FindOrCreateLmStateIndexForHistory(const std::vector<int32> &hist);
+
+  // Finds and returns the most specific LM-state index for a history or
+  // backed-off versions of it, that exists and has nonzero count.  Will die if
+  // there is no such history.  [e.g. if there is no unigram backoff state,
+  // which generally speaking there won't be.]
+  int32 FindNonzeroLmStateIndexForHistory(std::vector<int32> hist) const;
+
+  // after all backoff has been done, assigns FST state indexes to all states
+  // that exist and have nonzero count.  Returns the number of states.
+  int32 AssignFstStates();
+
+  // find the FST index of the initial-state, and returns it.
+  int32 FindInitialFstState() const;
+
+  // Write to an FST
+  void OutputToFst(
+      int32 num_fst_states,
+      fst::StdVectorFst *fst) const;
+
+};
+
+#endif
+

src/model.cc

@@ -14,19 +14,7 @@
 
 
 //
-// Possible model layout:
-//
-// * Default kaldi model with HCLG.fst
-//
-// * Lookahead model with const G.fst
-//
-// * Lookahead model with ngram G.fst
-//
-// * File disambig_tid.int required only for lookadhead models
-//
-// * File word_boundary.int is required if we want to have precise word timing information
-//   otherwise we don't do any word alignment. Optionally lexicon alignment can be done
-//   with corresponding C++ code inside kaldi recognizer.
+// For details of possible model layout see doc/models.md section model-structure
 
 #include "model.h"
 
@@ -45,24 +33,101 @@ static FstRegisterer<NGramFst<StdArc>> NGramFst_StdArc_registerer;
 
 #ifdef __ANDROID__
 #include <android/log.h>
-static void AndroidLogHandler(const LogMessageEnvelope &env, const char *message)
+static void KaldiLogHandler(const LogMessageEnvelope &env, const char *message)
 {
-    __android_log_print(ANDROID_LOG_VERBOSE, "KaldiDemo", message, 1);
+  int priority;
+  if (env.severity > GetVerboseLevel())
+      return;
+
+  if (env.severity > LogMessageEnvelope::kInfo) {
+    priority = ANDROID_LOG_VERBOSE;
+  } else {
+    switch (env.severity) {
+    case LogMessageEnvelope::kInfo:
+      priority = ANDROID_LOG_INFO;
+      break;
+    case LogMessageEnvelope::kWarning:
+      priority = ANDROID_LOG_WARN;
+      break;
+    case LogMessageEnvelope::kAssertFailed:
+      priority = ANDROID_LOG_FATAL;
+      break;
+    case LogMessageEnvelope::kError:
+    default: // If not the ERROR, it still an error!
+      priority = ANDROID_LOG_ERROR;
+      break;
+    }
+  }
+
+  std::stringstream full_message;
+  full_message << env.func << "():" << env.file << ':'
+               << env.line << ") " << message;
+
+  __android_log_print(priority, "VoskAPI", "%s", full_message.str().c_str());
+}
+#else
+static void KaldiLogHandler(const LogMessageEnvelope &env, const char *message)
+{
+  if (env.severity > GetVerboseLevel())
+      return;
+
+  // Modified default Kaldi logging so we can disable LOG messages.
+  std::stringstream full_message;
+  if (env.severity > LogMessageEnvelope::kInfo) {
+    full_message << "VLOG[" << env.severity << "] (";
+  } else {
+    switch (env.severity) {
+    case LogMessageEnvelope::kInfo:
+      full_message << "LOG (";
+      break;
+    case LogMessageEnvelope::kWarning:
+      full_message << "WARNING (";
+      break;
+    case LogMessageEnvelope::kAssertFailed:
+      full_message << "ASSERTION_FAILED (";
+      break;
+    case LogMessageEnvelope::kError:
+    default: // If not the ERROR, it still an error!
+      full_message << "ERROR (";
+      break;
+    }
+  }
+  // Add other info from the envelope and the message text.
+  full_message << "VoskAPI" << ':'
+               << env.func << "():" << env.file << ':'
+               << env.line << ") " << message;
+
+  // Print the complete message to stderr.
+  full_message << "\n";
+  std::cerr << full_message.str();
 }
 #endif
 
-Model::Model(const char *model_path) {
+Model::Model(const char *model_path) : model_path_str_(model_path) {
 
-#ifdef __ANDROID__
-    SetLogHandler(AndroidLogHandler);
-#endif
+    SetLogHandler(KaldiLogHandler);
 
-    const char *usage = "Read the docs";
+    struct stat buffer;
+    string am_path = model_path_str_ + "/am/final.mdl";
+    if (stat(am_path.c_str(), &buffer) == 0) {
+         ConfigureV2();
+    } else {
+         ConfigureV1();
+    }
+
+    ReadDataFiles();
+
+    ref_cnt_ = 1;
+}
+
+// Old model layout without model configuration file
+
+void Model::ConfigureV1()
+{
     const char *extra_args[] = {
-        "--min-active=200",
-        "--max-active=3000",
-        "--beam=10.0",
-        "--lattice-beam=2.0",
+        "--max-active=7000",
+        "--beam=13.0",
+        "--lattice-beam=6.0",
         "--acoustic-scale=1.0",
 
         "--frame-subsampling-factor=3",
@@ -71,49 +136,70 @@ Model::Model(const char *model_path) {
         "--endpoint.rule2.min-trailing-silence=0.5",
         "--endpoint.rule3.min-trailing-silence=1.0",
         "--endpoint.rule4.min-trailing-silence=2.0",
-    };
-    std::string model_path_str(model_path);
 
-    kaldi::ParseOptions po(usage);
+        "--print-args=false",
+    };
+
+    kaldi::ParseOptions po("");
     nnet3_decoding_config_.Register(&po);
     endpoint_config_.Register(&po);
     decodable_opts_.Register(&po);
 
-    std::vector<const char*> args;
+    vector<const char*> args;
     args.push_back("vosk");
     args.insert(args.end(), extra_args, extra_args + sizeof(extra_args) / sizeof(extra_args[0]));
     po.Read(args.size(), args.data());
 
+    nnet3_rxfilename_ = model_path_str_ + "/final.mdl";
+    hclg_fst_rxfilename_ = model_path_str_ + "/HCLG.fst";
+    hcl_fst_rxfilename_ = model_path_str_ + "/HCLr.fst";
+    g_fst_rxfilename_ = model_path_str_ + "/Gr.fst";
+    disambig_rxfilename_ = model_path_str_ + "/disambig_tid.int";
+    word_syms_rxfilename_ = model_path_str_ + "/words.txt";
+    winfo_rxfilename_ = model_path_str_ + "/word_boundary.int";
+    carpa_rxfilename_ = model_path_str_ + "/rescore/G.carpa";
+    std_fst_rxfilename_ = model_path_str_ + "/rescore/G.fst";
+    final_ie_rxfilename_ = model_path_str_ + "/ivector/final.ie";
+    mfcc_conf_rxfilename_ = model_path_str_ + "/mfcc.conf";
+}
+
+void Model::ConfigureV2()
+{
+    kaldi::ParseOptions po("something");
+    nnet3_decoding_config_.Register(&po);
+    endpoint_config_.Register(&po);
+    decodable_opts_.Register(&po);
+    po.ReadConfigFile(model_path_str_ + "/conf/model.conf");
+
+
+    nnet3_rxfilename_ = model_path_str_ + "/am/final.mdl";
+    hclg_fst_rxfilename_ = model_path_str_ + "/graph/HCLG.fst";
+    hcl_fst_rxfilename_ = model_path_str_ + "/graph/HCLr.fst";
+    g_fst_rxfilename_ = model_path_str_ + "/graph/Gr.fst";
+    disambig_rxfilename_ = model_path_str_ + "/graph/disambig_tid.int";
+    word_syms_rxfilename_ = model_path_str_ + "/graph/words.txt";
+    winfo_rxfilename_ = model_path_str_ + "/graph/phones/word_boundary.int";
+    carpa_rxfilename_ = model_path_str_ + "/rescore/G.carpa";
+    std_fst_rxfilename_ = model_path_str_ + "/rescore/G.fst";
+    final_ie_rxfilename_ = model_path_str_ + "/ivector/final.ie";
+    mfcc_conf_rxfilename_ = model_path_str_ + "/conf/mfcc.conf";
+}
+
+void Model::ReadDataFiles()
+{
+    struct stat buffer;
+
+    KALDI_LOG << "Decoding params beam=" << nnet3_decoding_config_.beam <<
+         " max-active=" << nnet3_decoding_config_.max_active <<
+         " lattice-beam=" << nnet3_decoding_config_.lattice_beam;
+    KALDI_LOG << "Silence phones " << endpoint_config_.silence_phones;
+
     feature_info_.feature_type = "mfcc";
-    ReadConfigFromFile(model_path_str + "/mfcc.conf", &feature_info_.mfcc_opts);
+    ReadConfigFromFile(mfcc_conf_rxfilename_, &feature_info_.mfcc_opts);
     feature_info_.mfcc_opts.frame_opts.allow_downsample = true; // It is safe to downsample
 
     feature_info_.silence_weighting_config.silence_weight = 1e-3;
-    feature_info_.silence_weighting_config.silence_phones_str = "1:2:3:4:5:6:7:8:9:10";
-
-    OnlineIvectorExtractionConfig ivector_extraction_opts;
-    ivector_extraction_opts.splice_config_rxfilename = model_path_str + "/ivector/splice.conf";
-    ivector_extraction_opts.cmvn_config_rxfilename = model_path_str + "/ivector/online_cmvn.conf";
-    ivector_extraction_opts.lda_mat_rxfilename = model_path_str + "/ivector/final.mat";
-    ivector_extraction_opts.global_cmvn_stats_rxfilename = model_path_str + "/ivector/global_cmvn.stats";
-    ivector_extraction_opts.diag_ubm_rxfilename = model_path_str + "/ivector/final.dubm";
-    ivector_extraction_opts.ivector_extractor_rxfilename = model_path_str + "/ivector/final.ie";
-    ivector_extraction_opts.num_gselect = 5;
-    ivector_extraction_opts.min_post = 0.025;
-    ivector_extraction_opts.posterior_scale = 0.1;
-    ivector_extraction_opts.max_remembered_frames = 1000;
-    ivector_extraction_opts.max_count = 100;
-    ivector_extraction_opts.ivector_period = 200;
-    feature_info_.use_ivectors = true;
-    feature_info_.ivector_extractor_info.Init(ivector_extraction_opts);
-
-    nnet3_rxfilename_ = model_path_str + "/final.mdl";
-    hclg_fst_rxfilename_ = model_path_str + "/HCLG.fst";
-    hcl_fst_rxfilename_ = model_path_str + "/HCLr.fst";
-    g_fst_rxfilename_ = model_path_str + "/Gr.fst";
-    disambig_rxfilename_ = model_path_str + "/disambig_tid.int";
-    word_syms_rxfilename_ = model_path_str + "/words.txt";
-    winfo_rxfilename_ = model_path_str + "/word_boundary.int";
+    feature_info_.silence_weighting_config.silence_phones_str = endpoint_config_.silence_phones;
 
     trans_model_ = new kaldi::TransitionModel();
     nnet_ = new kaldi::nnet3::AmNnetSimple();
@@ -126,19 +212,37 @@ Model::Model(const char *model_path) {
         SetDropoutTestMode(true, &(nnet_->GetNnet()));
         nnet3::CollapseModel(nnet3::CollapseModelConfig(), &(nnet_->GetNnet()));
     }
-
     decodable_info_ = new nnet3::DecodableNnetSimpleLoopedInfo(decodable_opts_,
                                                                nnet_);
-    struct stat buffer;
+
+    if (stat(final_ie_rxfilename_.c_str(), &buffer) == 0) {
+        KALDI_LOG << "Loading i-vector extractor from " << final_ie_rxfilename_;
+
+        OnlineIvectorExtractionConfig ivector_extraction_opts;
+        ivector_extraction_opts.splice_config_rxfilename = model_path_str_ + "/ivector/splice.conf";
+        ivector_extraction_opts.cmvn_config_rxfilename = model_path_str_ + "/ivector/online_cmvn.conf";
+        ivector_extraction_opts.lda_mat_rxfilename = model_path_str_ + "/ivector/final.mat";
+        ivector_extraction_opts.global_cmvn_stats_rxfilename = model_path_str_ + "/ivector/global_cmvn.stats";
+        ivector_extraction_opts.diag_ubm_rxfilename = model_path_str_ + "/ivector/final.dubm";
+        ivector_extraction_opts.ivector_extractor_rxfilename = model_path_str_ + "/ivector/final.ie";
+        ivector_extraction_opts.max_count = 100;
+
+        feature_info_.use_ivectors = true;
+        feature_info_.ivector_extractor_info.Init(ivector_extraction_opts);
+    } else {
+        feature_info_.use_ivectors = false;
+    }
+
     if (stat(hclg_fst_rxfilename_.c_str(), &buffer) == 0) {
+        KALDI_LOG << "Loading HCLG from " << hclg_fst_rxfilename_;
         hclg_fst_ = fst::ReadFstKaldiGeneric(hclg_fst_rxfilename_);
         hcl_fst_ = NULL;
         g_fst_ = NULL;
     } else {
+        KALDI_LOG << "Loading HCL and G from " << hcl_fst_rxfilename_ << " " << g_fst_rxfilename_;
         hclg_fst_ = NULL;
         hcl_fst_ = fst::StdFst::Read(hcl_fst_rxfilename_);
         g_fst_ = fst::StdFst::Read(g_fst_rxfilename_);
-
         ReadIntegerVectorSimple(disambig_rxfilename_, &disambig_);
     }
 
@@ -149,6 +253,7 @@ Model::Model(const char *model_path) {
         word_syms_ = g_fst_->OutputSymbols();
     }
     if (!word_syms_) {
+        KALDI_LOG << "Loading words from " << word_syms_rxfilename_;
         if (!(word_syms_ = fst::SymbolTable::ReadText(word_syms_rxfilename_)))
             KALDI_ERR << "Could not read symbol table from file "
                       << word_syms_rxfilename_;
@@ -156,11 +261,38 @@ Model::Model(const char *model_path) {
     KALDI_ASSERT(word_syms_);
 
     if (stat(winfo_rxfilename_.c_str(), &buffer) == 0) {
+        KALDI_LOG << "Loading winfo " << winfo_rxfilename_;
         kaldi::WordBoundaryInfoNewOpts opts;
         winfo_ = new kaldi::WordBoundaryInfo(opts, winfo_rxfilename_);
     } else {
         winfo_ = NULL;
     }
+
+    if (stat(carpa_rxfilename_.c_str(), &buffer) == 0) {
+        KALDI_LOG << "Loading CARPA model from " << carpa_rxfilename_;
+        std_lm_fst_ = fst::ReadFstKaldi(std_fst_rxfilename_);
+        fst::Project(std_lm_fst_, fst::PROJECT_OUTPUT);
+        if (std_lm_fst_->Properties(fst::kILabelSorted, true) == 0) {
+            fst::ILabelCompare<fst::StdArc> ilabel_comp;
+            fst::ArcSort(std_lm_fst_, ilabel_comp);
+        }
+        ReadKaldiObject(carpa_rxfilename_, &const_arpa_);
+    } else {
+        std_lm_fst_ = NULL;
+    }
+}
+
+void Model::Ref() 
+{
+    ref_cnt_++;
+}
+
+void Model::Unref() 
+{
+    ref_cnt_--;
+    if (ref_cnt_ == 0) {
+        delete this;
+    }
 }
 
 Model::~Model() {

src/model.h

@@ -12,8 +12,8 @@
 // See the License for the specific language governing permissions and
 // limitations under the License.
 
-#ifndef MODEL_H_
-#define MODEL_H_
+#ifndef VOSK_MODEL_H
+#define VOSK_MODEL_H
 
 #include "base/kaldi-common.h"
 #include "fstext/fstext-lib.h"
@@ -32,6 +32,7 @@
 #include "rnnlm/rnnlm-utils.h"
 
 using namespace kaldi;
+using namespace std;
 
 class KaldiRecognizer;
 
@@ -39,18 +40,29 @@ class Model {
 
 public:
     Model(const char *model_path);
-    ~Model();
+    void Ref();
+    void Unref();
 
 protected:
+    ~Model();
+    void ConfigureV1();
+    void ConfigureV2();
+    void ReadDataFiles();
+
     friend class KaldiRecognizer;
 
-    std::string nnet3_rxfilename_;
-    std::string hclg_fst_rxfilename_;
-    std::string hcl_fst_rxfilename_;
-    std::string g_fst_rxfilename_;
-    std::string word_syms_rxfilename_;
-    std::string winfo_rxfilename_;
-    std::string disambig_rxfilename_;
+    string model_path_str_;
+    string nnet3_rxfilename_;
+    string hclg_fst_rxfilename_;
+    string hcl_fst_rxfilename_;
+    string g_fst_rxfilename_;
+    string disambig_rxfilename_;
+    string word_syms_rxfilename_;
+    string winfo_rxfilename_;
+    string carpa_rxfilename_;
+    string std_fst_rxfilename_;
+    string final_ie_rxfilename_;
+    string mfcc_conf_rxfilename_;
 
     kaldi::OnlineEndpointConfig endpoint_config_;
     kaldi::LatticeFasterDecoderConfig nnet3_decoding_config_;
@@ -62,11 +74,16 @@ protected:
     kaldi::nnet3::AmNnetSimple *nnet_;
     const fst::SymbolTable *word_syms_;
     kaldi::WordBoundaryInfo *winfo_;
-    std::vector<int32> disambig_;
+    vector<int32> disambig_;
 
     fst::Fst<fst::StdArc> *hclg_fst_;
     fst::Fst<fst::StdArc> *hcl_fst_;
     fst::Fst<fst::StdArc> *g_fst_;
+
+    fst::VectorFst<fst::StdArc> *std_lm_fst_;
+    kaldi::ConstArpaLm const_arpa_;
+
+    int ref_cnt_;
 };
 
-#endif /* MODEL_H_ */
+#endif /* VOSK_MODEL_H */

src/spk_model.cc

@@ -24,4 +24,22 @@ SpkModel::SpkModel(const char *speaker_path) {
     SetBatchnormTestMode(true, &speaker_nnet);
     SetDropoutTestMode(true, &speaker_nnet);
     CollapseModel(nnet3::CollapseModelConfig(), &speaker_nnet);
+
+    ReadKaldiObject(speaker_path_str + "/mean.vec", &mean);
+    ReadKaldiObject(speaker_path_str + "/transform.mat", &transform);
+
+    ref_cnt_ = 1;
+}
+
+void SpkModel::Ref() 
+{
+    ref_cnt_++;
+}
+
+void SpkModel::Unref() 
+{
+    ref_cnt_--;
+    if (ref_cnt_ == 0) {
+        delete this;
+    }
 }

src/spk_model.h

@@ -12,8 +12,8 @@
 // See the License for the specific language governing permissions and
 // limitations under the License.
 
-#ifndef SPK_MODEL_H_
-#define SPK_MODEL_H_
+#ifndef VOSK_SPK_MODEL_H
+#define VOSK_SPK_MODEL_H
 
 #include "base/kaldi-common.h"
 #include "online2/online-feature-pipeline.h"
@@ -27,12 +27,20 @@ class SpkModel {
 
 public:
     SpkModel(const char *spk_path);
+    void Ref();
+    void Unref();
 
 protected:
     friend class KaldiRecognizer;
+    ~SpkModel() {};
 
     kaldi::nnet3::Nnet speaker_nnet;
+    kaldi::Vector<BaseFloat> mean;
+    kaldi::Matrix<BaseFloat> transform;
+
     MfccOptions spkvector_mfcc_opts;
+
+    int ref_cnt_;
 };
 
-#endif /* SPK_MODEL_H_ */
+#endif /* VOSK_SPK_MODEL_H */

src/vosk.i

@@ -1,7 +1,10 @@
-%module(package="vosk") vosk
+#if SWIGPYTHON
+%module(package="vosk", "threads"=1) vosk
+#else
+%module Vosk
+#endif
 
 %include <typemaps.i>
-%include <std_string.i>
 
 #if SWIGPYTHON
 %include <pybuffer.i>
@@ -11,39 +14,12 @@
 %include <arrays_csharp.i>
 #endif
 
-namespace kaldi {
-}
-
 #if SWIGPYTHON
 %pybuffer_binary(const char *data, int len);
-%ignore KaldiRecognizer::AcceptWaveform(const short *sdata, int len);
-%ignore KaldiRecognizer::AcceptWaveform(const float *fdata, int len);
-%exception {
-  try {
-    $action
-  } catch (kaldi::KaldiFatalError &e) {
-    PyErr_SetString(PyExc_RuntimeError, const_cast<char*>(e.KaldiMessage()));
-    SWIG_fail;
-  } catch (std::exception &e) {
-    PyErr_SetString(PyExc_RuntimeError, const_cast<char*>(e.what()));
-    SWIG_fail;
-  }
-}
 #endif
 
 #if SWIGJAVA
 %apply char *BYTE {const char *data};
-%ignore KaldiRecognizer::AcceptWaveform(const short *sdata, int len);
-%ignore KaldiRecognizer::AcceptWaveform(const float *fdata, int len);
-#endif
-
-%{
-#include "kaldi_recognizer.h"
-#include "model.h"
-#include "spk_model.h"
-%}
-
-#if SWIGJAVA
 %typemap(javaimports) KaldiRecognizer %{
 import java.nio.ByteBuffer;
 import java.nio.ByteOrder;
@@ -58,6 +34,11 @@ import java.nio.ByteOrder;
     return AcceptWaveform(bdata, bdata.length);
   }
 %}
+%pragma(java) jniclasscode=%{
+    static {
+        System.loadLibrary("vosk_jni");
+    }
+%}
 #endif
 
 #if SWIGCSHARP
@@ -67,6 +48,94 @@ CSHARP_ARRAYS(char, byte)
 %apply short INPUT[] {const short *sdata};
 #endif
 
-%include "kaldi_recognizer.h"
-%include "model.h"
-%include "spk_model.h"
+
+#if SWIGJAVASCRIPT
+%begin %{
+#include <v8.h>
+#include <node.h>
+#include <node_buffer.h>
+%}
+#endif
+
+%{
+#include "vosk_api.h"
+typedef struct VoskModel Model;
+typedef struct VoskSpkModel SpkModel;
+typedef struct VoskRecognizer KaldiRecognizer;
+%}
+
+typedef struct {} Model;
+typedef struct {} SpkModel;
+typedef struct {} KaldiRecognizer;
+
+%extend Model {
+    Model(const char *model_path)  {
+        return vosk_model_new(model_path);
+    }
+    ~Model() {
+        vosk_model_free($self);
+    }
+}
+
+%extend SpkModel {
+    SpkModel(const char *model_path)  {
+        return vosk_spk_model_new(model_path);
+    }
+    ~SpkModel() {
+        vosk_spk_model_free($self);
+    }
+}
+
+%extend KaldiRecognizer {
+    KaldiRecognizer(Model *model, float sample_rate)  {
+        return vosk_recognizer_new(model, sample_rate);
+    }
+    KaldiRecognizer(Model *model, SpkModel *spk_model, float sample_rate)  {
+        return vosk_recognizer_new_spk(model, spk_model, sample_rate);
+    }
+    KaldiRecognizer(Model *model, float sample_rate, const char* grammar)  {
+        return vosk_recognizer_new_grm(model, sample_rate, grammar);
+    }
+    ~KaldiRecognizer() {
+        vosk_recognizer_free($self);
+    }
+
+#if SWIGCSHARP
+    bool AcceptWaveform(const char *data, int len) {
+        return vosk_recognizer_accept_waveform($self, data, len);
+    }
+    bool AcceptWaveform(const short *sdata, int len) {
+        return vosk_recognizer_accept_waveform_s($self, sdata, len);
+    }
+    bool AcceptWaveform(const float *fdata, int len) {
+        return vosk_recognizer_accept_waveform_f($self, fdata, len);
+    }
+#elif SWIGJAVA
+    bool AcceptWaveform(const char *data, int len) {
+        return vosk_recognizer_accept_waveform($self, data, len);
+    }
+#elif SWIGJAVASCRIPT
+    bool AcceptWaveform(SWIG_Object ptr) {
+        char* data = (char*) node::Buffer::Data(ptr);
+        size_t length = node::Buffer::Length(ptr);
+        return vosk_recognizer_accept_waveform($self, data, length);
+    }
+#else
+    int AcceptWaveform(const char *data, int len) {
+        return vosk_recognizer_accept_waveform($self, data, len);
+    }
+#endif
+
+    const char* Result() {
+        return vosk_recognizer_result($self);
+    }
+    const char* PartialResult() {
+        return vosk_recognizer_partial_result($self);
+    }
+    const char* FinalResult() {
+        return vosk_recognizer_final_result($self);
+    }
+}
+
+%rename(SetLogLevel) vosk_set_log_level;
+void vosk_set_log_level(int level);

src/vosk_api.cc

@@ -0,0 +1,97 @@
+// Copyright 2020 Alpha Cephei Inc.
+//
+// Licensed under the Apache License, Version 2.0 (the "License");
+// you may not use this file except in compliance with the License.
+// You may obtain a copy of the License at
+//
+//       http://www.apache.org/licenses/LICENSE-2.0
+//
+// Unless required by applicable law or agreed to in writing, software
+// distributed under the License is distributed on an "AS IS" BASIS,
+// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+// See the License for the specific language governing permissions and
+// limitations under the License.
+
+#include "vosk_api.h"
+#include "kaldi_recognizer.h"
+#include "model.h"
+#include "spk_model.h"
+
+#include <string.h>
+
+using namespace kaldi;
+
+VoskModel *vosk_model_new(const char *model_path)
+{
+    return (VoskModel *)new Model(model_path);
+}
+
+void vosk_model_free(VoskModel *model)
+{
+    ((Model *)model)->Unref();
+}
+
+VoskSpkModel *vosk_spk_model_new(const char *model_path)
+{
+    return (VoskSpkModel *)new SpkModel(model_path);
+}
+
+void vosk_spk_model_free(VoskSpkModel *model)
+{
+    ((SpkModel *)model)->Unref();
+}
+
+VoskRecognizer *vosk_recognizer_new(VoskModel *model, float sample_rate)
+{
+    return (VoskRecognizer *)new KaldiRecognizer((Model *)model, sample_rate);
+}
+
+VoskRecognizer *vosk_recognizer_new_spk(VoskModel *model, VoskSpkModel *spk_model, float sample_rate)
+{
+    return (VoskRecognizer *)new KaldiRecognizer((Model *)model, (SpkModel *)spk_model, sample_rate);
+}
+
+VoskRecognizer *vosk_recognizer_new_grm(VoskModel *model, float sample_rate, const char *grammar)
+{
+    return (VoskRecognizer *)new KaldiRecognizer((Model *)model, sample_rate, grammar);
+}
+
+int vosk_recognizer_accept_waveform(VoskRecognizer *recognizer, const char *data, int length)
+{
+    return ((KaldiRecognizer *)(recognizer))->AcceptWaveform(data, length);
+}
+
+int vosk_recognizer_accept_waveform_s(VoskRecognizer *recognizer, const short *data, int length)
+{
+    return ((KaldiRecognizer *)(recognizer))->AcceptWaveform(data, length);
+}
+
+int vosk_recognizer_accept_waveform_f(VoskRecognizer *recognizer, const float *data, int length)
+{
+    return ((KaldiRecognizer *)(recognizer))->AcceptWaveform(data, length);
+}
+
+const char *vosk_recognizer_result(VoskRecognizer *recognizer)
+{
+    return ((KaldiRecognizer *)recognizer)->Result();
+}
+
+const char *vosk_recognizer_partial_result(VoskRecognizer *recognizer)
+{
+    return ((KaldiRecognizer *)recognizer)->PartialResult();
+}
+
+const char *vosk_recognizer_final_result(VoskRecognizer *recognizer)
+{
+    return ((KaldiRecognizer *)recognizer)->FinalResult();
+}
+
+void vosk_recognizer_free(VoskRecognizer *recognizer)
+{
+    delete (KaldiRecognizer *)(recognizer);
+}
+
+void vosk_set_log_level(int log_level)
+{
+    SetVerboseLevel(log_level);
+}

src/vosk_api.h

@@ -0,0 +1,211 @@
+// Copyright 2020 Alpha Cephei Inc.
+//
+// Licensed under the Apache License, Version 2.0 (the "License");
+// you may not use this file except in compliance with the License.
+// You may obtain a copy of the License at
+//
+//       http://www.apache.org/licenses/LICENSE-2.0
+//
+// Unless required by applicable law or agreed to in writing, software
+// distributed under the License is distributed on an "AS IS" BASIS,
+// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+// See the License for the specific language governing permissions and
+// limitations under the License.
+
+/* This header contains the C API for Vosk speech recognition system */
+
+#ifndef VOSK_API_H
+#define VOSK_API_H
+
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+/** Model stores all the data required for recognition
+ *  it contains static data and can be shared across processing
+ *  threads. */
+typedef struct VoskModel VoskModel;
+
+
+/** Speaker model is the same as model but contains the data
+ *  for speaker identification. */
+typedef struct VoskSpkModel VoskSpkModel;
+
+
+/** Recognizer object is the main object which processes data.
+ *  Each recognizer usually runs in own thread and takes audio as input.
+ *  Once audio is processed recognizer returns JSON object as a string
+ *  which represent decoded information - words, confidences, times, n-best lists,
+ *  speaker information and so on */
+typedef struct VoskRecognizer VoskRecognizer;
+
+
+/** Loads model data from the file and returns the model object
+ *
+ * @param model_path: the path of the model on the filesystem
+ @ @returns model object */
+VoskModel *vosk_model_new(const char *model_path);
+
+
+/** Releases the model memory
+ *
+ *  The model object is reference-counted so if some recognizer
+ *  depends on this model, model might still stay alive. When
+ *  last recognizer is released, model will be released too. */
+void vosk_model_free(VoskModel *model);
+
+
+/** Loads speaker model data from the file and returns the model object
+ *
+ * @param model_path: the path of the model on the filesystem
+ * @returns model object */
+VoskSpkModel *vosk_spk_model_new(const char *model_path);
+
+
+/** Releases the model memory
+ *
+ *  The model object is reference-counted so if some recognizer
+ *  depends on this model, model might still stay alive. When
+ *  last recognizer is released, model will be released too. */
+void vosk_spk_model_free(VoskSpkModel *model);
+
+/** Creates the recognizer object
+ *
+ *  The recognizers process the speech and return text using shared model data 
+ *  @param sample_rate The sample rate of the audio you going to feed into the recognizer
+ *  @returns recognizer object */
+VoskRecognizer *vosk_recognizer_new(VoskModel *model, float sample_rate);
+
+
+/** Creates the recognizer object with speaker recognition
+ *
+ *  With the speaker recognition mode the recognizer not just recognize
+ *  text but also return speaker vectors one can use for speaker identification
+ *
+ *  @param spk_model speaker model for speaker identification
+ *  @param sample_rate The sample rate of the audio you going to feed into the recognizer
+ *  @returns recognizer object */
+VoskRecognizer *vosk_recognizer_new_spk(VoskModel *model, VoskSpkModel *spk_model, float sample_rate);
+
+
+/** Creates the recognizer object with the phrase list
+ *
+ *  Sometimes when you want to improve recognition accuracy and when you don't need
+ *  to recognize large vocabulary you can specify a list of phrases to recognize. This
+ *  will improve recognizer speed and accuracy but might return [unk] if user said
+ *  something different.
+ *
+ *  Only recognizers with lookahead models support this type of quick configuration.
+ *  Precompiled HCLG graph models are not supported.
+ *
+ *  @param sample_rate The sample rate of the audio you going to feed into the recognizer
+ *  @param grammar The string with the list of phrases to recognize as JSON array of strings,
+ *                 for example "["one two three four five", "[unk]"]".
+ *
+ *  @returns recognizer object */
+VoskRecognizer *vosk_recognizer_new_grm(VoskModel *model, float sample_rate, const char *grammar);
+
+
+/** Accept voice data
+ *
+ *  accept and process new chunk of voice data
+ *
+ *  @param data - audio data in PCM 16-bit mono format
+ *  @param length - length of the audio data
+ *  @returns true if silence is occured and you can retrieve a new utterance with result method */
+int vosk_recognizer_accept_waveform(VoskRecognizer *recognizer, const char *data, int length);
+
+
+/** Same as above but the version with the short data for language bindings where you have
+ *  audio as array of shorts */
+int vosk_recognizer_accept_waveform_s(VoskRecognizer *recognizer, const short *data, int length);
+
+
+/** Same as above but the version with the float data for language bindings where you have
+ *  audio as array of floats */
+int vosk_recognizer_accept_waveform_f(VoskRecognizer *recognizer, const float *data, int length);
+
+
+/** Returns speech recognition result
+ *
+ * @returns the result in JSON format which contains decoded line, decoded
+ *          words, times in seconds and confidences. You can parse this result
+ *          with any json parser
+ *
+ * <pre>
+ * {
+ *   "result" : [{
+ *       "conf" : 1.000000,
+ *       "end" : 1.110000,
+ *       "start" : 0.870000,
+ *       "word" : "what"
+ *     }, {
+ *       "conf" : 1.000000,
+ *       "end" : 1.530000,
+ *       "start" : 1.110000,
+ *       "word" : "zero"
+ *     }, {
+ *       "conf" : 1.000000,
+ *       "end" : 1.950000,
+ *       "start" : 1.530000,
+ *       "word" : "zero"
+ *     }, {
+ *       "conf" : 1.000000,
+ *       "end" : 2.340000,
+ *       "start" : 1.950000,
+ *       "word" : "zero"
+ *     }, {
+ *       "conf" : 1.000000,
+ *      "end" : 2.610000,
+ *       "start" : 2.340000,
+ *       "word" : "one"
+ *     }],
+ *   "text" : "what zero zero zero one"
+ *  }
+ * </pre>
+ */
+const char *vosk_recognizer_result(VoskRecognizer *recognizer);
+
+
+/** Returns partial speech recognition
+ *
+ * @returns partial speech recognition text which is not yet finalized.
+ *          result may change as recognizer process more data.
+ *
+ * <pre>
+ * {
+ *  "partial" : "cyril one eight zero"
+ * }
+ * </pre>
+ */
+const char *vosk_recognizer_partial_result(VoskRecognizer *recognizer);
+
+
+/** Returns speech recognition result. Same as result, but doesn't wait for silence
+ *  You usually call it in the end of the stream to get final bits of audio. It
+ *  flushes the feature pipeline, so all remaining audio chunks got processed.
+ *
+ *  @returns speech result in JSON format.
+ */
+const char *vosk_recognizer_final_result(VoskRecognizer *recognizer);
+
+
+/** Releases recognizer object
+ *
+ *  Underlying model is also unreferenced and if needed released */
+void vosk_recognizer_free(VoskRecognizer *recognizer);
+
+/** Set log level for Kaldi messages
+ *
+ *  @param log_level the level
+ *     0 - default value to print info and error messages but no debug
+ *     less than 0 - don't print info messages
+ *     greather than 0 - more verbose mode
+ */
+void vosk_set_log_level(int log_level);
+
+#ifdef __cplusplus
+}
+#endif
+
+#endif /* VOSK_API_H */

travis/Dockerfile.dockcross

@@ -10,6 +10,9 @@ RUN apt-get update && \
         libffi-dev \
         libpcre3-dev \
         zlib1g-dev \
+        automake \
+        autoconf \
+        libtool \
     && rm -rf /var/lib/apt/lists/*
 
 RUN cd /opt \
@@ -20,21 +23,6 @@ RUN cd /opt \
     && cd .. \
     && rm -rf swig-4.0.1.tar.gz swig-4.0.1
 
-ARG OPENBLAS_ARCH=ARMV7
-RUN cd /opt \
-    && export OPENFST_CONFIGURE="--enable-static --enable-shared --enable-far --enable-ngram-fsts --enable-lookahead-fsts --with-pic --disable-bin --host=${CROSS_TRIPLE} --build=x86-linux-gnu" \
-    && git clone -b lookahead --single-branch https://github.com/alphacep/kaldi \
-    && cd kaldi/tools \
-    && git clone -b v0.3.7 --single-branch https://github.com/xianyi/OpenBLAS \
-    && make PREFIX=$(pwd)/OpenBLAS/install TARGET="${OPENBLAS_ARCH}" HOSTCC=gcc USE_LOCKING=1 USE_THREAD=0 -C OpenBLAS all install \
-    && sed -i 's:status=0:exit 0:g' extras/check_dependencies.sh \
-    && make -j 10 openfst \
-    && cd /opt/kaldi/src \
-    && sed -i "s:TARGET_ARCH=\"\`uname -m\`\":TARGET_ARCH=$(echo $CROSS_TRIPLE|cut -d - -f 1):g" configure \
-    && ./configure --mathlib=OPENBLAS --shared --use-cuda=no \
-    && make -j 10 online2 \
-    && find /opt/kaldi -name "*.o" -exec rm {} \;
-
 RUN cd /opt \
     && wget -q https://github.com/openssl/openssl/archive/OpenSSL_1_0_2u.tar.gz \
     && tar xf OpenSSL_1_0_2u.tar.gz \
@@ -60,3 +48,57 @@ RUN cd /opt \
     && make -j $(nproc) \
     && make install \
     && rm -rf /opt/cpython-3.7.6 /opt/cpython-3.7.6-cross /opt/v3.7.6.tar.gz
+
+RUN cd /opt \
+    && wget -q https://github.com/python/cpython/archive/v3.6.10.tar.gz \
+    && tar xf v3.6.10.tar.gz \
+    && cp -r cpython-3.6.10 cpython-3.6.10-cross \
+    && cd /opt/cpython-3.6.10 \
+    && AR=/usr/bin/ar RANLIB=/usr/bin/ranlib CPP=/usr/bin/cpp CXX=/usr/bin/g++ CC=/usr/bin/gcc  ./configure --prefix="/opt/python/cp3.6-cp3.6m" \
+    && make -j $(nproc) \
+    && make install \
+    && /opt/python/cp3.6-cp3.6m/bin/pip3 install -U pip \
+    && /opt/python/cp3.6-cp3.6m/bin/pip3 install -U wheel \
+    && cd /opt/cpython-3.6.10-cross \
+    && export PATH=/opt/python/cp3.6-cp3.6m/bin:$PATH \
+    && ./configure --prefix=$CROSS_ROOT --with-openssl=$CROSS_ROOT --host=${CROSS_TRIPLE} --build=x86-linux-gnu --disable-ipv6 ac_cv_file__dev_ptmx=no ac_cv_file__dev_ptc=no ac_cv_have_long_long_format=yes \
+    && make -j $(nproc) \
+    && make install \
+    && rm -rf /opt/cpython-3.6.10 /opt/cpython-3.6.10-cross /opt/v3.6.10.tar.gz
+
+RUN cd /opt \
+    && wget -q https://github.com/python/cpython/archive/v3.8.3.tar.gz \
+    && tar xf v3.8.3.tar.gz \
+    && cp -r cpython-3.8.3 cpython-3.8.3-cross \
+    && cd /opt/cpython-3.8.3 \
+    && AR=/usr/bin/ar RANLIB=/usr/bin/ranlib CPP=/usr/bin/cpp CXX=/usr/bin/g++ CC=/usr/bin/gcc  ./configure --prefix="/opt/python/cp3.8-cp3.8m" \
+    && make -j $(nproc) \
+    && make install \
+    && /opt/python/cp3.8-cp3.8m/bin/pip3 install -U pip \
+    && /opt/python/cp3.8-cp3.8m/bin/pip3 install -U wheel \
+    && cd /opt/cpython-3.8.3-cross \
+    && export PATH=/opt/python/cp3.8-cp3.8m/bin:$PATH \
+    && ./configure --prefix=$CROSS_ROOT --with-openssl=$CROSS_ROOT --host=${CROSS_TRIPLE} --build=x86-linux-gnu --disable-ipv6 ac_cv_file__dev_ptmx=no ac_cv_file__dev_ptc=no ac_cv_have_long_long_format=yes \
+    && make -j $(nproc) \
+    && make install \
+    && rm -rf /opt/cpython-3.8.3 /opt/cpython-3.8.3-cross /opt/v3.8.3.tar.gz
+
+ARG OPENBLAS_ARCH=ARMV7
+ARG ARM_HARDWARE_OPTS="-mfloat-abi=hard -mfpu=neon"
+RUN cd /opt \
+    && git clone -b lookahead --single-branch https://github.com/alphacep/kaldi \
+    && cd kaldi/tools \
+    && git clone -b v0.3.7 --single-branch https://github.com/xianyi/OpenBLAS \
+    && make PREFIX=$(pwd)/OpenBLAS/install TARGET="${OPENBLAS_ARCH}" HOSTCC=gcc USE_LOCKING=1 USE_THREAD=0 -C OpenBLAS all install \
+    && git clone -b old-gcc --single-branch https://github.com/alphacep/openfst openfst \
+    && cd openfst \
+    && autoreconf -i \
+    && ./configure --prefix=/opt/kaldi/tools/openfst --enable-static --enable-shared --enable-far --enable-ngram-fsts --enable-lookahead-fsts --with-pic --disable-bin --host=${CROSS_TRIPLE} --build=x86-linux-gnu \
+    && make -j 10 && make install \
+    && cd /opt/kaldi/src \
+    && sed -i "s:TARGET_ARCH=\"\`uname -m\`\":TARGET_ARCH=$(echo $CROSS_TRIPLE|cut -d - -f 1):g" configure \
+    && sed -i "s:-mfloat-abi=hard -mfpu=neon:${ARM_HARDWARE_OPTS}:g" makefiles/linux_openblas_arm.mk \
+    && sed -i "s: -O1 : -O3 :g" makefiles/linux_openblas_arm.mk \
+    && ./configure --mathlib=OPENBLAS --shared --use-cuda=no \
+    && make -j 10 online2 lm \
+    && find /opt/kaldi -name "*.o" -exec rm {} \;

travis/Dockerfile.manylinux

@@ -7,29 +7,11 @@ RUN yum -y update && yum -y install \
     wget \
     openssl-devel \
     pcre-devel \
+    automake \
+    autoconf \
+    libtool \
     && yum clean all
 
-RUN cd /opt \
-    && wget https://github.com/Kitware/CMake/releases/download/v3.16.2/cmake-3.16.2.tar.gz \
-    && tar xf cmake-3.16.2.tar.gz \
-    && cd cmake-3.16.2 \
-    && ./configure --prefix=/usr && make -j 10 && make install \
-    && cd .. \
-    && rm -rf cmake-3.16.2 cmake-3.16.2.tar.gz
-
-RUN cd /opt \
-    && export OPENFST_CONFIGURE="--enable-static --enable-shared --enable-far --enable-ngram-fsts --enable-lookahead-fsts --with-pic --disable-bin" \
-    && git clone -b lookahead --single-branch https://github.com/alphacep/kaldi \
-    && cd kaldi/tools \
-    && git clone -b v0.3.7 --single-branch https://github.com/xianyi/OpenBLAS \
-    && make PREFIX=$(pwd)/OpenBLAS/install TARGET=NEHALEM USE_LOCKING=1 USE_THREAD=0 -C OpenBLAS all install \
-    && sed -i 's:status=0:exit 0:g' extras/check_dependencies.sh \
-    && make -j 10 openfst \
-    && cd ../src \
-    && ./configure --mathlib=OPENBLAS --shared --use-cuda=no \
-    && make -j 10 online2 \
-    && find /opt/kaldi -name "*.o" -exec rm {} \;
-
 RUN cd /opt \
     && wget -O swig-4.0.1.tar.gz https://sourceforge.net/projects/swig/files/swig/swig-4.0.1/swig-4.0.1.tar.gz/download \
     && tar xf swig-4.0.1.tar.gz \
@@ -38,3 +20,17 @@ RUN cd /opt \
     && cd .. \
     && rm -rf swig-4.0.1.tar.gz swig-4.0.1
 
+RUN cd /opt \
+    && git clone -b lookahead --single-branch https://github.com/alphacep/kaldi \
+    && cd kaldi/tools \
+    && git clone -b v0.3.7 --single-branch https://github.com/xianyi/OpenBLAS \
+    && make PREFIX=$(pwd)/OpenBLAS/install DYNAMIC_ARCH=1 USE_LOCKING=1 USE_THREAD=0 -C OpenBLAS all install \
+    && git clone https://github.com/alphacep/openfst openfst \
+    && cd openfst \
+    && autoreconf -i \
+    && ./configure --prefix=/opt/kaldi/tools/openfst --enable-static --enable-shared --enable-far --enable-ngram-fsts --enable-lookahead-fsts --with-pic --disable-bin \
+    && make -j 10 && make install \
+    && cd /opt/kaldi/src \
+    && ./configure --mathlib=OPENBLAS --shared --use-cuda=no \
+    && make -j 10 online2 lm rnnlm \
+    && find /opt/kaldi -name "*.o" -exec rm {} \;

travis/build-dockcross.sh

@@ -4,5 +4,9 @@ set -e
 set -x
 
 docker build --build-arg="DOCKCROSS_IMAGE=linux-armv7" --build-arg="OPENBLAS_ARCH=ARMV7" --file Dockerfile.dockcross --tag alphacep/kaldi-dockcross-armv7:latest .
-docker build --build-arg="DOCKCROSS_IMAGE=linux-armv6" --build-arg="OPENBLAS_ARCH=ARMV6" --file Dockerfile.dockcross --tag alphacep/kaldi-dockcross-armv6:latest .
+docker build --build-arg="DOCKCROSS_IMAGE=linux-armv6" --build-arg="OPENBLAS_ARCH=ARMV6" --build-arg="ARM_HARDWARE_OPTS=" --file Dockerfile.dockcross --tag alphacep/kaldi-dockcross-armv6:latest .
 docker build --build-arg="DOCKCROSS_IMAGE=linux-arm64" --build-arg="OPENBLAS_ARCH=ARMV8" --file Dockerfile.dockcross --tag alphacep/kaldi-dockcross-arm64:latest .
+
+docker run --rm -v /home/shmyrev/travis/vosk-api/:/io alphacep/kaldi-dockcross-armv6 /io/travis/build-wheels-dockcross.sh
+docker run --rm -v /home/shmyrev/travis/vosk-api/:/io alphacep/kaldi-dockcross-armv7 /io/travis/build-wheels-dockcross.sh
+docker run --rm -v /home/shmyrev/travis/vosk-api/:/io alphacep/kaldi-dockcross-arm64 /io/travis/build-wheels-dockcross.sh

travis/build-docker.sh

@@ -4,3 +4,4 @@ set -e
 set -x
 
 docker build --file Dockerfile.manylinux --tag alphacep/kaldi-manylinux:latest .
+docker run --rm -e PLAT=manylinux2010_x86_64 -v /home/shmyrev/travis/vosk-api/:/io alphacep/kaldi-manylinux /io/travis/build-wheels.sh

travis/build-wheels-dockcross.sh

@@ -1,20 +1,37 @@
 #!/bin/bash
 set -e -x
 
-export KALDI_ROOT=/opt/kaldi
-export WHEEL_FLAGS=`$CROSS_ROOT/bin/python3-config --cflags`
-export PATH=/opt/python/cp3.7-cp3.7m/bin:$PATH
-echo $CROSS_TRIPLE
-case $CROSS_TRIPLE in
-    *arm-*)
-        export _PYTHON_HOST_PLATFORM=linux-armv6l
-        ;;
-    *armv7-*)
-        export _PYTHON_HOST_PLATFORM=linux-armv7l
-        ;;
-    *aarch64-*)
-        export _PYTHON_HOST_PLATFORM=linux-aarch64
-        ;;
-esac
+ORIG_PATH=$PATH
+for pyver in 3.6 3.7 3.8; do
 
-pip3 wheel /io/python -w /io/wheelhouse
+    export KALDI_ROOT=/opt/kaldi
+    export PATH=/opt/python/cp${pyver}-cp${pyver}m/bin:$ORIG_PATH
+    export VOSK_SOURCE=/io/src
+
+    # Python 3.8 somehow changed syconfig file name
+    sysconfig_bit="m"
+    if [ $pyver == "3.8" ]; then
+         sysconfig_bit=""
+    fi
+
+    case $CROSS_TRIPLE in
+        *arm-*)
+            export _PYTHON_HOST_PLATFORM=linux-armv6l
+            export _PYTHON_SYSCONFIGDATA_NAME=_sysconfigdata_${sysconfig_bit}_linux_arm-linux-gnueabihf
+            ;;
+        *armv7-*)
+            export _PYTHON_HOST_PLATFORM=linux-armv7l
+            export _PYTHON_SYSCONFIGDATA_NAME=_sysconfigdata_${sysconfig_bit}_linux_arm-linux-gnueabihf
+            ;;
+        *aarch64-*)
+            export _PYTHON_HOST_PLATFORM=linux-aarch64
+            export _PYTHON_SYSCONFIGDATA_NAME=_sysconfigdata_${sysconfig_bit}_linux_aarch64-linux-gnu
+            ;;
+    esac
+    export PYTHONHOME=$CROSS_ROOT
+    export PYTHONPATH=/opt/python/cp${pyver}-cp${pyver}m/lib/python${pyver}/site-packages:/opt/python/cp${pyver}-cp${pyver}m/lib/python${pyver}/lib-dynload
+
+    rm -rf /io/python/build
+    pip${pyver} -v wheel /io/python -w /io/wheelhouse
+
+done

travis/build-wheels.sh

@@ -4,10 +4,11 @@ set -e -x
 export KALDI_ROOT=/opt/kaldi
 
 # Compile wheels
-for pypath in /opt/python/cp3*; do
-    export WHEEL_FLAGS=`${pypath}/bin/python3-config --cflags`
+for pypath in /opt/python/cp3[56789]*; do
+    export VOSK_SOURCE=/io/src
     mkdir -p /opt/wheelhouse
-    "${pypath}/bin/pip" wheel /io/python -w /opt/wheelhouse
+    rm -rf /io/python/build
+    "${pypath}/bin/pip" -v wheel /io/python -w /opt/wheelhouse
 done
 
 # Bundle external shared libraries into the wheels

webjs/index.js

@@ -0,0 +1,3 @@
+exports.printMsg = function() {
+  console.log("This is a message from the Vosk package");
+}

webjs/package.json

@@ -0,0 +1,20 @@
+{
+  "name": "vosk-js",
+  "version": "0.3.0",
+  "description": "Node binding for continuous voice recoginition through vosk-api.",
+  "repository": {
+    "type": "git",
+    "url": "git://github.com/alphacep/vosk-api.git"
+  },
+  "main": "index.js",
+  "keywords": [
+    "speech",
+    "speech recognition",
+    "voice"
+  ],
+  "author": "Alpha Cephei Inc.",
+  "license": "Apache 2.0",
+  "engines": {
+    "node": ">= 12.x.x"
+  }
+}