Add Kotlin and Java API for Moonshine models (#1474)
This commit is contained in:
Fangjun Kuang
60
java-api-examples/NonStreamingDecodeFileMoonshine.java
Normal file
60
java-api-examples/NonStreamingDecodeFileMoonshine.java
Normal file
@@ -0,0 +1,60 @@
|
||||
// Copyright 2024 Xiaomi Corporation
|
||||
|
||||
// This file shows how to use an offline Moonshine,
|
||||
// i.e., non-streaming Moonshine model,
|
||||
// to decode files.
|
||||
import com.k2fsa.sherpa.onnx.*;
|
||||
|
||||
public class NonStreamingDecodeFileMoonshine {
|
||||
public static void main(String[] args) {
|
||||
// please refer to
|
||||
// https://k2-fsa.github.io/sherpa/onnx/moonshine/index.html
|
||||
// to download model files
|
||||
|
||||
String preprocessor = "./sherpa-onnx-moonshine-tiny-en-int8/preprocess.onnx";
|
||||
String encoder = "./sherpa-onnx-moonshine-tiny-en-int8/encode.int8.onnx";
|
||||
String uncachedDecoder = "./sherpa-onnx-moonshine-tiny-en-int8/uncached_decode.int8.onnx";
|
||||
String cachedDecoder = "./sherpa-onnx-moonshine-tiny-en-int8/cached_decode.int8.onnx";
|
||||
|
||||
String tokens = "./sherpa-onnx-moonshine-tiny-en-int8/tokens.txt";
|
||||
|
||||
String waveFilename = "./sherpa-onnx-moonshine-tiny-en-int8/test_wavs/0.wav";
|
||||
|
||||
WaveReader reader = new WaveReader(waveFilename);
|
||||
|
||||
OfflineMoonshineModelConfig moonshine =
|
||||
OfflineMoonshineModelConfig.builder()
|
||||
.setPreprocessor(preprocessor)
|
||||
.setEncoder(encoder)
|
||||
.setUncachedDecoder(uncachedDecoder)
|
||||
.setCachedDecoder(cachedDecoder)
|
||||
.build();
|
||||
|
||||
OfflineModelConfig modelConfig =
|
||||
OfflineModelConfig.builder()
|
||||
.setMoonshine(moonshine)
|
||||
.setTokens(tokens)
|
||||
.setNumThreads(1)
|
||||
.setDebug(true)
|
||||
.build();
|
||||
|
||||
OfflineRecognizerConfig config =
|
||||
OfflineRecognizerConfig.builder()
|
||||
.setOfflineModelConfig(modelConfig)
|
||||
.setDecodingMethod("greedy_search")
|
||||
.build();
|
||||
|
||||
OfflineRecognizer recognizer = new OfflineRecognizer(config);
|
||||
OfflineStream stream = recognizer.createStream();
|
||||
stream.acceptWaveform(reader.getSamples(), reader.getSampleRate());
|
||||
|
||||
recognizer.decode(stream);
|
||||
|
||||
String text = recognizer.getResult(stream).getText();
|
||||
|
||||
System.out.printf("filename:%s\nresult:%s\n", waveFilename, text);
|
||||
|
||||
stream.release();
|
||||
recognizer.release();
|
||||
}
|
||||
}
|
||||
152
java-api-examples/VadFromMicWithNonStreamingMoonshine.java
Normal file
152
java-api-examples/VadFromMicWithNonStreamingMoonshine.java
Normal file
@@ -0,0 +1,152 @@
|
||||
// Copyright 2024 Xiaomi Corporation
|
||||
|
||||
// This file shows how to use a silero_vad model with a non-streaming
|
||||
// Moonshine tiny for speech recognition.
|
||||
|
||||
import com.k2fsa.sherpa.onnx.*;
|
||||
import javax.sound.sampled.*;
|
||||
|
||||
public class VadFromMicNonStreamingMoonshine {
|
||||
private static final int sampleRate = 16000;
|
||||
private static final int windowSize = 512;
|
||||
|
||||
public static Vad createVad() {
|
||||
// please download ./silero_vad.onnx from
|
||||
// https://github.com/k2-fsa/sherpa-onnx/releases/tag/asr-models
|
||||
String model = "./silero_vad.onnx";
|
||||
SileroVadModelConfig sileroVad =
|
||||
SileroVadModelConfig.builder()
|
||||
.setModel(model)
|
||||
.setThreshold(0.5f)
|
||||
.setMinSilenceDuration(0.25f)
|
||||
.setMinSpeechDuration(0.5f)
|
||||
.setWindowSize(windowSize)
|
||||
.build();
|
||||
|
||||
VadModelConfig config =
|
||||
VadModelConfig.builder()
|
||||
.setSileroVadModelConfig(sileroVad)
|
||||
.setSampleRate(sampleRate)
|
||||
.setNumThreads(1)
|
||||
.setDebug(true)
|
||||
.setProvider("cpu")
|
||||
.build();
|
||||
|
||||
return new Vad(config);
|
||||
}
|
||||
|
||||
public static OfflineRecognizer createOfflineRecognizer() {
|
||||
// please refer to
|
||||
// https://k2-fsa.github.io/sherpa/onnx/moonshine/index.html
|
||||
// to download model files
|
||||
|
||||
String preprocessor = "./sherpa-onnx-moonshine-tiny-en-int8/preprocess.onnx";
|
||||
String encoder = "./sherpa-onnx-moonshine-tiny-en-int8/encode.int8.onnx";
|
||||
String uncachedDecoder = "./sherpa-onnx-moonshine-tiny-en-int8/uncached_decode.int8.onnx";
|
||||
String cachedDecoder = "./sherpa-onnx-moonshine-tiny-en-int8/cached_decode.int8.onnx";
|
||||
|
||||
String tokens = "./sherpa-onnx-moonshine-tiny-en-int8/tokens.txt";
|
||||
|
||||
OfflineMoonshineModelConfig moonshine =
|
||||
OfflineMoonshineModelConfig.builder()
|
||||
.setPreprocessor(preprocessor)
|
||||
.setEncoder(encoder)
|
||||
.setUncachedDecoder(uncachedDecoder)
|
||||
.setCachedDecoder(cachedDecoder)
|
||||
.build();
|
||||
|
||||
OfflineModelConfig modelConfig =
|
||||
OfflineModelConfig.builder()
|
||||
.setMoonshine(moonshine)
|
||||
.setTokens(tokens)
|
||||
.setNumThreads(1)
|
||||
.setDebug(true)
|
||||
.build();
|
||||
|
||||
OfflineRecognizerConfig config =
|
||||
OfflineRecognizerConfig.builder()
|
||||
.setOfflineModelConfig(modelConfig)
|
||||
.setDecodingMethod("greedy_search")
|
||||
.build();
|
||||
|
||||
return new OfflineRecognizer(config);
|
||||
}
|
||||
|
||||
public static void main(String[] args) {
|
||||
Vad vad = createVad();
|
||||
OfflineRecognizer recognizer = createOfflineRecognizer();
|
||||
|
||||
// https://docs.oracle.com/javase/8/docs/api/javax/sound/sampled/AudioFormat.html
|
||||
// Linear PCM, 16000Hz, 16-bit, 1 channel, signed, little endian
|
||||
AudioFormat format = new AudioFormat(sampleRate, 16, 1, true, false);
|
||||
|
||||
// https://docs.oracle.com/javase/8/docs/api/javax/sound/sampled/DataLine.Info.html#Info-java.lang.Class-javax.sound.sampled.AudioFormat-int-
|
||||
DataLine.Info info = new DataLine.Info(TargetDataLine.class, format);
|
||||
TargetDataLine targetDataLine;
|
||||
try {
|
||||
targetDataLine = (TargetDataLine) AudioSystem.getLine(info);
|
||||
targetDataLine.open(format);
|
||||
targetDataLine.start();
|
||||
} catch (LineUnavailableException e) {
|
||||
System.out.println("Failed to open target data line: " + e.getMessage());
|
||||
vad.release();
|
||||
recognizer.release();
|
||||
return;
|
||||
}
|
||||
|
||||
boolean printed = false;
|
||||
byte[] buffer = new byte[windowSize * 2];
|
||||
float[] samples = new float[windowSize];
|
||||
|
||||
System.out.println("Started. Please speak");
|
||||
boolean running = true;
|
||||
while (targetDataLine.isOpen() && running) {
|
||||
int n = targetDataLine.read(buffer, 0, buffer.length);
|
||||
if (n <= 0) {
|
||||
System.out.printf("Got %d bytes. Expected %d bytes.\n", n, buffer.length);
|
||||
continue;
|
||||
}
|
||||
for (int i = 0; i != windowSize; ++i) {
|
||||
short low = buffer[2 * i];
|
||||
short high = buffer[2 * i + 1];
|
||||
int s = (high << 8) + low;
|
||||
samples[i] = (float) s / 32768;
|
||||
}
|
||||
|
||||
vad.acceptWaveform(samples);
|
||||
if (vad.isSpeechDetected() && !printed) {
|
||||
System.out.println("Detected speech");
|
||||
printed = true;
|
||||
}
|
||||
|
||||
if (!vad.isSpeechDetected()) {
|
||||
printed = false;
|
||||
}
|
||||
|
||||
while (!vad.empty()) {
|
||||
SpeechSegment segment = vad.front();
|
||||
float startTime = segment.getStart() / (float) sampleRate;
|
||||
float duration = segment.getSamples().length / (float) sampleRate;
|
||||
|
||||
OfflineStream stream = recognizer.createStream();
|
||||
stream.acceptWaveform(segment.getSamples(), sampleRate);
|
||||
recognizer.decode(stream);
|
||||
String text = recognizer.getResult(stream).getText();
|
||||
stream.release();
|
||||
|
||||
if (!text.isEmpty()) {
|
||||
System.out.printf("%.3f--%.3f: %s\n", startTime, startTime + duration, text);
|
||||
}
|
||||
|
||||
if (text.contains("exit the program")) {
|
||||
running = false;
|
||||
}
|
||||
|
||||
vad.pop();
|
||||
}
|
||||
}
|
||||
|
||||
vad.release();
|
||||
recognizer.release();
|
||||
}
|
||||
}
|
||||
37
java-api-examples/run-non-streaming-decode-file-moonshine.sh
Executable file
37
java-api-examples/run-non-streaming-decode-file-moonshine.sh
Executable file
@@ -0,0 +1,37 @@
|
||||
#!/usr/bin/env bash
|
||||
|
||||
set -ex
|
||||
|
||||
if [[ ! -f ../build/lib/libsherpa-onnx-jni.dylib && ! -f ../build/lib/libsherpa-onnx-jni.so ]]; then
|
||||
mkdir -p ../build
|
||||
pushd ../build
|
||||
cmake \
|
||||
-DSHERPA_ONNX_ENABLE_PYTHON=OFF \
|
||||
-DSHERPA_ONNX_ENABLE_TESTS=OFF \
|
||||
-DSHERPA_ONNX_ENABLE_CHECK=OFF \
|
||||
-DBUILD_SHARED_LIBS=ON \
|
||||
-DSHERPA_ONNX_ENABLE_PORTAUDIO=OFF \
|
||||
-DSHERPA_ONNX_ENABLE_JNI=ON \
|
||||
..
|
||||
|
||||
make -j4
|
||||
ls -lh lib
|
||||
popd
|
||||
fi
|
||||
|
||||
if [ ! -f ../sherpa-onnx/java-api/build/sherpa-onnx.jar ]; then
|
||||
pushd ../sherpa-onnx/java-api
|
||||
make
|
||||
popd
|
||||
fi
|
||||
|
||||
if [ ! -f ./sherpa-onnx-moonshine-tiny-en-int8/tokens.txt ]; then
|
||||
curl -SL -O https://github.com/k2-fsa/sherpa-onnx/releases/download/asr-models/sherpa-onnx-moonshine-tiny-en-int8.tar.bz2
|
||||
tar xvf sherpa-onnx-moonshine-tiny-en-int8.tar.bz2
|
||||
rm sherpa-onnx-moonshine-tiny-en-int8.tar.bz2
|
||||
fi
|
||||
|
||||
java \
|
||||
-Djava.library.path=$PWD/../build/lib \
|
||||
-cp ../sherpa-onnx/java-api/build/sherpa-onnx.jar \
|
||||
NonStreamingDecodeFileMoonshine.java
|
||||
41
java-api-examples/run-vad-from-mic-non-streaming-moonshine.sh
Executable file
41
java-api-examples/run-vad-from-mic-non-streaming-moonshine.sh
Executable file
@@ -0,0 +1,41 @@
|
||||
#!/usr/bin/env bash
|
||||
|
||||
set -ex
|
||||
|
||||
if [[ ! -f ../build/lib/libsherpa-onnx-jni.dylib && ! -f ../build/lib/libsherpa-onnx-jni.so ]]; then
|
||||
mkdir -p ../build
|
||||
pushd ../build
|
||||
cmake \
|
||||
-DSHERPA_ONNX_ENABLE_PYTHON=OFF \
|
||||
-DSHERPA_ONNX_ENABLE_TESTS=OFF \
|
||||
-DSHERPA_ONNX_ENABLE_CHECK=OFF \
|
||||
-DBUILD_SHARED_LIBS=ON \
|
||||
-DSHERPA_ONNX_ENABLE_PORTAUDIO=OFF \
|
||||
-DSHERPA_ONNX_ENABLE_JNI=ON \
|
||||
..
|
||||
|
||||
make -j4
|
||||
ls -lh lib
|
||||
popd
|
||||
fi
|
||||
|
||||
if [ ! -f ../sherpa-onnx/java-api/build/sherpa-onnx.jar ]; then
|
||||
pushd ../sherpa-onnx/java-api
|
||||
make
|
||||
popd
|
||||
fi
|
||||
|
||||
if [ ! -f ./silero_vad.onnx ]; then
|
||||
curl -SL -O https://github.com/k2-fsa/sherpa-onnx/releases/download/asr-models/silero_vad.onnx
|
||||
fi
|
||||
|
||||
if [ ! -f ./sherpa-onnx-moonshine-tiny-en-int8/tokens.txt ]; then
|
||||
curl -SL -O https://github.com/k2-fsa/sherpa-onnx/releases/download/asr-models/sherpa-onnx-moonshine-tiny-en-int8.tar.bz2
|
||||
tar xvf sherpa-onnx-moonshine-tiny-en-int8.tar.bz2
|
||||
rm sherpa-onnx-moonshine-tiny-en-int8.tar.bz2
|
||||
fi
|
||||
|
||||
java \
|
||||
-Djava.library.path=$PWD/../build/lib \
|
||||
-cp ../sherpa-onnx/java-api/build/sherpa-onnx.jar \
|
||||
./VadFromMicWithNonStreamingMoonshine.java
|
||||
Reference in New Issue
Block a user