【通用视觉】集成 OpenAI CLIP 模型，支持以图搜图、以文搜图、以图搜文等功能

【通用视觉】新增 YOLO 图像分类模型支持
【ASR/TTS】集成 Sherpa TTS（语音合成）与 ASR（语音识别）模块，支持中文、粤语、方言、英文等多种语言
【目标检测】优化视频目标检测功能

examples/speech-examples/pom.xml

@@ -12,7 +12,7 @@
         <maven.compiler.source>11</maven.compiler.source>
         <maven.compiler.target>11</maven.compiler.target>
         <project.build.sourceEncoding>UTF-8</project.build.sourceEncoding>
-        <smartjavaai.version>1.0.25</smartjavaai.version>
+        <smartjavaai.version>1.0.26</smartjavaai.version>
         <!--如果打包运行，需要替换成你的main-->
         <exec.mainClass>smartai.examples.speech.asr.common.OcrRecognizeDemo</exec.mainClass>
 

examples/speech-examples/src/main/java/smartai/examples/speech/asr/SherpaAsrDemo.java

@@ -0,0 +1,389 @@
+package smartai.examples.speech.asr;
+
+import ai.djl.modality.audio.Audio;
+import ai.djl.modality.audio.AudioFactory;
+import ai.djl.util.JsonUtils;
+import cn.smartjavaai.common.entity.Language;
+import cn.smartjavaai.common.entity.R;
+import cn.smartjavaai.speech.asr.config.AsrModelConfig;
+import cn.smartjavaai.speech.asr.entity.AsrResult;
+import cn.smartjavaai.speech.asr.entity.WhisperParams;
+import cn.smartjavaai.speech.asr.enums.AsrModelEnum;
+import cn.smartjavaai.speech.asr.factory.SpeechRecognizerFactory;
+import cn.smartjavaai.speech.asr.model.SpeechRecognizer;
+import lombok.extern.slf4j.Slf4j;
+import org.junit.Test;
+
+import java.nio.file.Paths;
+
+/**
+ * 语音识别ASR demo
+ * sherpa-onnx模型及依赖库下载链接:
+ * 1、（推荐）依赖库官网下载：https://github.com/k2-fsa/sherpa-onnx/releases
+ * 2、（推荐）ASR模型官网下载：https://github.com/k2-fsa/sherpa-onnx/releases/tag/asr-models
+ * 3、百度网盘下载：https://pan.baidu.com/s/19p3WhVEM7dgdkvXFaeeAxg?pwd=1234 提取码: 1234
+ * @author dwj
+ * @date 2025/10/23
+ */
+@Slf4j
+public class SherpaAsrDemo {
+
+
+    /**
+     * 语音识别：Sherpa Paraformer（中文）
+     */
+    @Test
+    public void testSherpaParaformerAsr() {
+        try {
+            //获取模型
+            AsrModelConfig config = new AsrModelConfig();
+            config.setModelEnum(AsrModelEnum.SHERPA_PARAFORMER);
+            config.setModelPath("/Users/wenjie/Documents/develop/model/speech/sherpa-asr/sherpa-onnx-paraformer-zh-2023-09-14");
+            config.setModelName("model.int8.onnx");
+            config.setLibPath(Paths.get("/Users/wenjie/smartjavaai_cache/sherpa-onnx-v1.12.14-osx-arm64-jni/lib"));
+            config.putCustomParam("debug", false);
+            config.putCustomParam("numThreads", 1);
+            SpeechRecognizer recognizer = SpeechRecognizerFactory.getInstance().getModel(config);
+            Audio audio = AudioFactory.newInstance().fromFile(Paths.get("/Users/wenjie/Documents/develop/model/speech/sherpa-asr/sherpa-onnx-paraformer-zh-2023-09-14/test_wavs/0.wav"));
+            R<AsrResult> result = recognizer.recognize(audio);
+            if (result.isSuccess()){
+                log.info("识别成功:{}", result.getData());
+            }else{
+                log.error("识别失败:{}", result.getMessage());
+            }
+        } catch (Exception e) {
+            e.printStackTrace();
+        }
+    }
+
+    /**
+     * 语音识别：Sherpa Dolphin（中文）
+     */
+    @Test
+    public void testSherpaDolphinAsr() {
+        try {
+            //获取模型
+            AsrModelConfig config = new AsrModelConfig();
+            config.setModelEnum(AsrModelEnum.SHERPA_DOLPHIN);
+            config.setModelName("model.int8.onnx");
+            config.setModelPath("/Users/wenjie/Documents/develop/model/speech/sherpa-asr/sherpa-onnx-dolphin-base-ctc-multi-lang-int8-2025-04-02");
+            config.setLibPath(Paths.get("/Users/wenjie/smartjavaai_cache/sherpa-onnx-v1.12.14-osx-arm64-jni/lib"));
+            config.putCustomParam("debug", false);
+            config.putCustomParam("numThreads", 1);
+            SpeechRecognizer recognizer = SpeechRecognizerFactory.getInstance().getModel(config);
+            Audio audio = AudioFactory.newInstance().fromFile(Paths.get("/Users/wenjie/Documents/develop/model/speech/sherpa-asr/sherpa-onnx-dolphin-base-ctc-multi-lang-int8-2025-04-02/test_wavs/0.wav"));
+            R<AsrResult> result = recognizer.recognize(audio);
+            if (result.isSuccess()){
+                log.info("识别成功:{}", result.getData());
+            }else{
+                log.error("识别失败:{}", result.getMessage());
+            }
+        } catch (Exception e) {
+            e.printStackTrace();
+        }
+    }
+
+    /**
+     * 语音识别：Sherpa zipformer（中文）
+     */
+    @Test
+    public void testSherpaZipformerAsr() {
+        try {
+            //获取模型
+            AsrModelConfig config = new AsrModelConfig();
+            config.setModelEnum(AsrModelEnum.SHERPA_ZIPFORMERCTC);
+            config.setModelPath("/Users/wenjie/Documents/develop/model/speech/sherpa-asr/sherpa-onnx-zipformer-ctc-zh-int8-2025-07-03");
+            config.setLibPath(Paths.get("/Users/wenjie/smartjavaai_cache/sherpa-onnx-v1.12.14-osx-arm64-jni/lib"));
+            config.putCustomParam("debug", false);
+            config.putCustomParam("numThreads", 1);
+            config.setModelName("model.int8.onnx");
+            SpeechRecognizer recognizer = SpeechRecognizerFactory.getInstance().getModel(config);
+            Audio audio = AudioFactory.newInstance().fromFile(Paths.get("/Users/wenjie/Documents/develop/model/speech/sherpa-asr/sherpa-onnx-zipformer-ctc-zh-int8-2025-07-03/test_wavs/0.wav"));
+            R<AsrResult> result = recognizer.recognize(audio);
+            if (result.isSuccess()){
+                log.info("识别成功:{}", result.getData());
+            }else{
+                log.error("识别失败:{}", result.getMessage());
+            }
+        } catch (Exception e) {
+            e.printStackTrace();
+        }
+    }
+
+    /**
+     * 语音识别：Sherpa FireRedAsr（中英）
+     */
+    @Test
+    public void testSherpaFireRedAsr() {
+        try {
+            //获取模型
+            AsrModelConfig config = new AsrModelConfig();
+            config.setModelEnum(AsrModelEnum.SHERPA_FIREREDASR);
+            config.setModelPath("/Users/wenjie/Documents/develop/model/speech/sherpa-asr/sherpa-onnx-fire-red-asr-large-zh_en-2025-02-16");
+            config.setLibPath(Paths.get("/Users/wenjie/smartjavaai_cache/sherpa-onnx-v1.12.14-osx-arm64-jni/lib"));
+            config.putCustomParam("debug", false);
+            config.putCustomParam("numThreads", 1);
+            SpeechRecognizer recognizer = SpeechRecognizerFactory.getInstance().getModel(config);
+            Audio audio = AudioFactory.newInstance().fromFile(Paths.get("/Users/wenjie/Documents/develop/model/speech/sherpa-asr/sherpa-onnx-fire-red-asr-large-zh_en-2025-02-16/test_wavs/3.wav"));
+            R<AsrResult> result = recognizer.recognize(audio);
+            if (result.isSuccess()){
+                log.info("识别成功:{}", result.getData());
+            }else{
+                log.error("识别失败:{}", result.getMessage());
+            }
+        } catch (Exception e) {
+            e.printStackTrace();
+        }
+    }
+
+    /**
+     * 语音识别：Sherpa SenseVoice （多语言模型）
+     */
+    @Test
+    public void testSherpaSenseVoiceAsr() {
+        try {
+            //获取模型
+            AsrModelConfig config = new AsrModelConfig();
+            config.setModelEnum(AsrModelEnum.SHERPA_SENSEVOICE);
+            config.setModelPath("/Users/wenjie/Documents/develop/model/speech/sherpa-asr/sherpa-onnx-sense-voice-zh-en-ja-ko-yue-2024-07-17");
+            config.setModelName("model.onnx");
+            config.setLibPath(Paths.get("/Users/wenjie/smartjavaai_cache/sherpa-onnx-v1.12.14-osx-arm64-jni/lib"));
+            config.putCustomParam("debug", false);
+            config.putCustomParam("numThreads", 1);
+            SpeechRecognizer recognizer = SpeechRecognizerFactory.getInstance().getModel(config);
+            Audio audio = AudioFactory.newInstance().fromFile(Paths.get("/Users/wenjie/Documents/develop/model/speech/sherpa-asr/sherpa-onnx-sense-voice-zh-en-ja-ko-yue-2024-07-17/test_wavs/zh.wav"));
+            R<AsrResult> result = recognizer.recognize(audio);
+            if (result.isSuccess()){
+                log.info("识别成功:{}", result.getData());
+            }else{
+                log.error("识别失败:{}", result.getMessage());
+            }
+        } catch (Exception e) {
+            e.printStackTrace();
+        }
+    }
+
+    /**
+     * 语音识别：Sherpa WenetCtc（多语言模型：粤语）
+     */
+    @Test
+    public void testSherpaWenetCtcAsr() {
+        try {
+            //获取模型
+            AsrModelConfig config = new AsrModelConfig();
+            config.setModelEnum(AsrModelEnum.SHERPA_WENETCTC);
+            config.setModelPath("/Users/wenjie/Documents/develop/model/speech/sherpa-asr/sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10");
+            config.setModelName("model.int8.onnx");
+            config.setLibPath(Paths.get("/Users/wenjie/smartjavaai_cache/sherpa-onnx-v1.12.14-osx-arm64-jni/lib"));
+            config.putCustomParam("debug", false);
+            config.putCustomParam("numThreads", 1);
+            SpeechRecognizer recognizer = SpeechRecognizerFactory.getInstance().getModel(config);
+            Audio audio = AudioFactory.newInstance().fromFile(Paths.get("/Users/wenjie/Documents/develop/model/speech/sherpa-asr/sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/test_wavs/yue-0.wav"));
+            R<AsrResult> result = recognizer.recognize(audio);
+            if (result.isSuccess()){
+                log.info("识别成功:{}", result.getData());
+            }else{
+                log.error("识别失败:{}", result.getMessage());
+            }
+        } catch (Exception e) {
+            e.printStackTrace();
+        }
+    }
+
+    /**
+     * 语音识别：Sherpa FireRedAsr（方言：四川、天津、河南）
+     */
+    @Test
+    public void testSherpaFireRedAsrSichuan() {
+        try {
+            //获取模型
+            AsrModelConfig config = new AsrModelConfig();
+            config.setModelEnum(AsrModelEnum.SHERPA_FIREREDASR);
+            config.setModelPath("/Users/wenjie/Documents/develop/model/speech/sherpa-asr/sherpa-onnx-fire-red-asr-large-zh_en-2025-02-16");
+            config.setLibPath(Paths.get("/Users/wenjie/smartjavaai_cache/sherpa-onnx-v1.12.14-osx-arm64-jni/lib"));
+            config.putCustomParam("debug", false);
+            config.putCustomParam("numThreads", 1);
+            SpeechRecognizer recognizer = SpeechRecognizerFactory.getInstance().getModel(config);
+            Audio audio = AudioFactory.newInstance().fromFile(Paths.get("/Users/wenjie/Documents/develop/model/speech/sherpa-asr/sherpa-onnx-fire-red-asr-large-zh_en-2025-02-16/test_wavs/3-sichuan.wav"));
+            R<AsrResult> result = recognizer.recognize(audio);
+            if (result.isSuccess()){
+                log.info("识别成功:{}", result.getData());
+            }else{
+                log.error("识别失败:{}", result.getMessage());
+            }
+        } catch (Exception e) {
+            e.printStackTrace();
+        }
+    }
+
+    /**
+     * 语音识别：Sherpa Telespeech（方言-天津、河南、四川）
+     */
+    @Test
+    public void testSherpaTelespeechAsr() {
+        try {
+            //获取模型
+            AsrModelConfig config = new AsrModelConfig();
+            config.setModelEnum(AsrModelEnum.SHERPA_TELESPEECH);
+            config.setModelPath("/Users/wenjie/Documents/develop/model/speech/sherpa-asr/sherpa-onnx-telespeech-ctc-int8-zh-2024-06-04");
+            config.setLibPath(Paths.get("/Users/wenjie/smartjavaai_cache/sherpa-onnx-v1.12.14-osx-arm64-jni/lib"));
+            config.setModelName("model.int8.onnx");
+            config.putCustomParam("debug", false);
+            config.putCustomParam("numThreads", 1);
+            SpeechRecognizer recognizer = SpeechRecognizerFactory.getInstance().getModel(config);
+            Audio audio = AudioFactory.newInstance().fromFile(Paths.get("/Users/wenjie/Documents/develop/model/speech/sherpa-asr/sherpa-onnx-telespeech-ctc-int8-zh-2024-06-04/test_wavs/4-tianjin.wav"));
+            R<AsrResult> result = recognizer.recognize(audio);
+            if (result.isSuccess()){
+                log.info("识别成功:{}", result.getData());
+            }else{
+                log.error("识别失败:{}", result.getMessage());
+            }
+        } catch (Exception e) {
+            e.printStackTrace();
+        }
+    }
+
+    /**
+     * 语音识别：Sherpa Nemo（英文）
+     */
+    @Test
+    public void testSherpaNemoAsr() {
+        try {
+            //获取模型
+            AsrModelConfig config = new AsrModelConfig();
+            config.setModelEnum(AsrModelEnum.SHERPA_NEMO);
+            config.setModelPath("/Users/wenjie/Documents/develop/model/speech/sherpa-asr/sherpa-onnx-nemo-ctc-en-citrinet-512");
+            config.setModelName("model.onnx");
+            config.setLibPath(Paths.get("/Users/wenjie/smartjavaai_cache/sherpa-onnx-v1.12.14-osx-arm64-jni/lib"));
+            config.putCustomParam("debug", false);
+            config.putCustomParam("numThreads", 1);
+            SpeechRecognizer recognizer = SpeechRecognizerFactory.getInstance().getModel(config);
+            Audio audio = AudioFactory.newInstance().fromFile(Paths.get("/Users/wenjie/Documents/develop/model/speech/sherpa-asr/sherpa-onnx-nemo-ctc-en-citrinet-512/test_wavs/0.wav"));
+            R<AsrResult> result = recognizer.recognize(audio);
+            if (result.isSuccess()){
+                log.info("识别成功:{}", result.getData());
+            }else{
+                log.error("识别失败:{}", result.getMessage());
+            }
+        } catch (Exception e) {
+            e.printStackTrace();
+        }
+    }
+
+    /**
+     * 语音识别：Sherpa Moonshine（英文）
+     */
+    @Test
+    public void testSherpaMoonshineAsr() {
+        try {
+            //获取模型
+            AsrModelConfig config = new AsrModelConfig();
+            config.setModelEnum(AsrModelEnum.SHERPA_MOONSHINE);
+            config.setModelPath("/Users/wenjie/Documents/develop/model/speech/sherpa-asr/sherpa-onnx-moonshine-tiny-en-int8");
+            config.setLibPath(Paths.get("/Users/wenjie/smartjavaai_cache/sherpa-onnx-v1.12.14-osx-arm64-jni/lib"));
+            config.putCustomParam("debug", false);
+            config.putCustomParam("numThreads", 2);
+            SpeechRecognizer recognizer = SpeechRecognizerFactory.getInstance().getModel(config);
+            Audio audio = AudioFactory.newInstance().fromFile(Paths.get("/Users/wenjie/Documents/develop/model/speech/sherpa-asr/sherpa-onnx-moonshine-tiny-en-int8/test_wavs/0.wav"));
+            R<AsrResult> result = recognizer.recognize(audio);
+            if (result.isSuccess()){
+                log.info("识别成功:{}", result.getData());
+            }else{
+                log.error("识别失败:{}", result.getMessage());
+            }
+        } catch (Exception e) {
+            e.printStackTrace();
+        }
+    }
+
+
+
+    /**
+     * 语音识别：Sherpa Whisper（英文）
+     */
+    @Test
+    public void testSherpaWhisperAsr() {
+        try {
+            //获取模型
+            AsrModelConfig config = new AsrModelConfig();
+            config.setModelEnum(AsrModelEnum.SHERPA_WHISPER);
+            config.setModelPath("/Users/wenjie/Documents/develop/model/speech/sherpa-asr/sherpa-onnx-whisper-tiny");
+            config.setLibPath(Paths.get("/Users/wenjie/smartjavaai_cache/sherpa-onnx-v1.12.14-osx-arm64-jni/lib"));
+            config.putCustomParam("debug", false);
+            config.putCustomParam("numThreads", 1);
+            SpeechRecognizer recognizer = SpeechRecognizerFactory.getInstance().getModel(config);
+            Audio audio = AudioFactory.newInstance().fromFile(Paths.get("/Users/wenjie/Documents/develop/model/speech/sherpa-asr/sherpa-onnx-whisper-tiny/test_wavs/0.wav"));
+            R<AsrResult> result = recognizer.recognize(audio);
+            if (result.isSuccess()){
+                log.info("识别成功:{}", result.getData());
+            }else{
+                log.error("识别失败:{}", result.getMessage());
+            }
+        } catch (Exception e) {
+            e.printStackTrace();
+        }
+    }
+
+    /**
+     * 语音识别：Sherpa Transducer（英文）
+     */
+    @Test
+    public void testSherpaTransducerAsr() {
+        try {
+            //获取模型
+            AsrModelConfig config = new AsrModelConfig();
+            config.setModelEnum(AsrModelEnum.SHERPA_TRANSDUCER);
+            config.setModelPath("/Users/wenjie/Documents/develop/model/speech/sherpa-asr/sherpa-onnx-zipformer-gigaspeech-2023-12-12");
+            config.setLibPath(Paths.get("/Users/wenjie/smartjavaai_cache/sherpa-onnx-v1.12.14-osx-arm64-jni/lib"));
+            config.putCustomParam("debug", false);
+            config.putCustomParam("numThreads", 1);
+            SpeechRecognizer recognizer = SpeechRecognizerFactory.getInstance().getModel(config);
+            Audio audio = AudioFactory.newInstance().fromFile(Paths.get("/Users/wenjie/Documents/develop/model/speech/sherpa-asr/sherpa-onnx-zipformer-gigaspeech-2023-12-12/test_wavs/1221-135766-0001.wav"));
+            R<AsrResult> result = recognizer.recognize(audio);
+            if (result.isSuccess()){
+                log.info("识别成功:{}", result.getData());
+            }else{
+                log.error("识别失败:{}", result.getMessage());
+            }
+        } catch (Exception e) {
+            e.printStackTrace();
+        }
+    }
+
+    /**
+     * 语音识别：Sherpa Canary（德语-英文）
+     */
+    @Test
+    public void testSherpaCanaryAsr() {
+        try {
+            //获取模型
+            AsrModelConfig config = new AsrModelConfig();
+            config.setModelEnum(AsrModelEnum.SHERPA_CANARY);
+            config.setModelPath("/Users/wenjie/Documents/develop/model/speech/sherpa-asr/sherpa-onnx-nemo-canary-180m-flash-en-es-de-fr-int8");
+            config.setLibPath(Paths.get("/Users/wenjie/smartjavaai_cache/sherpa-onnx-v1.12.14-osx-arm64-jni/lib"));
+            config.putCustomParam("debug", false);
+            config.putCustomParam("numThreads", 1);
+            SpeechRecognizer recognizer = SpeechRecognizerFactory.getInstance().getModel(config);
+            Audio audio = AudioFactory.newInstance().fromFile(Paths.get("/Users/wenjie/Documents/develop/model/speech/sherpa-asr/sherpa-onnx-nemo-canary-180m-flash-en-es-de-fr-int8/test_wavs/de.wav"));
+            R<AsrResult> result = recognizer.recognize(audio);
+            if (result.isSuccess()){
+                log.info("识别成功:{}", result.getData());
+            }else{
+                log.error("识别失败:{}", result.getMessage());
+            }
+        } catch (Exception e) {
+            e.printStackTrace();
+        }
+    }
+
+
+
+
+
+
+
+
+
+
+
+}

examples/speech-examples/src/main/java/smartai/examples/speech/asr/SpeechRecognizeDemo.java

@@ -33,7 +33,7 @@ import java.io.InputStream;
 import java.nio.file.Paths;
 
 /**
- * 语音识别demo
+ * 语音识别demo（Vosk、Whisper）
  * 模型下载网盘：https://pan.baidu.com/s/1kiMF5MF641R7LTn1GpB2lQ?pwd=1234 提取码: 1234
  * 文档地址：http://doc.smartjavaai.cn/
  * @author dwj
@@ -41,6 +41,8 @@ import java.nio.file.Paths;
 @Slf4j
 public class SpeechRecognizeDemo {
 
+
+
     /**
      * 获取Whisper模型
      * 模型下载网盘：https://pan.baidu.com/s/1kiMF5MF641R7LTn1GpB2lQ?pwd=1234 提取码: 1234

examples/speech-examples/src/main/java/smartai/examples/speech/tts/TtsDemo.java

@@ -0,0 +1,193 @@
+package smartai.examples.speech.tts;
+
+
+import ai.djl.modality.audio.Audio;
+import cn.smartjavaai.common.entity.R;
+import cn.smartjavaai.common.enums.DeviceEnum;
+import cn.smartjavaai.speech.tts.config.TtsModelConfig;
+import cn.smartjavaai.speech.tts.entity.SherpaTtsParams;
+import cn.smartjavaai.speech.tts.enums.TtsModelEnum;
+import cn.smartjavaai.speech.tts.factory.TtsModelFactory;
+import cn.smartjavaai.speech.tts.model.SherpaTtsModel;
+import cn.smartjavaai.speech.tts.model.TtsModel;
+import cn.smartjavaai.speech.utils.AudioUtils;
+import lombok.extern.slf4j.Slf4j;
+import org.junit.Test;
+
+import java.io.IOException;
+import java.nio.file.Paths;
+
+/**
+ * 语音合成demo（TTS）
+ * sherpa-onnx模型及依赖库下载链接:
+ * 1、（推荐）依赖库官网下载：https://github.com/k2-fsa/sherpa-onnx/releases
+ * 2、（推荐）TTS模型官网下载：https://github.com/k2-fsa/sherpa-onnx/releases/tag/tts-models
+ * 3、百度网盘下载：https://pan.baidu.com/s/186REUf7p1z0HH9AZNnwCUg?pwd=1234 提取码: 1234
+ * @author dwj
+ * @date 2025/10/22
+ */
+@Slf4j
+public class TtsDemo {
+
+    public static String enText = "Today as always, men fall into two groups: slaves and free men. Whoever does not have"
+            + " two-thirds of his day for himself, is a slave, whatever he may be: a statesman, a"
+            + " businessman, an official, or a scholar.";
+
+    public static String zhText = "有问题，请拨打110或者手机18601239876。我们的价值观是真诚热爱！";
+
+    String znEnText =
+            "中英文语音合成测试。This is generated by next generation Kaldi using Kokoro without Misaki."
+                    + " 你觉得中英文说的如何呢？";
+
+    //设备类型
+    public static DeviceEnum device = DeviceEnum.CPU;
+
+
+    /**
+     * 获取Vits中文模型
+     * @return
+     */
+    public TtsModel getVitsZhModel(){
+        TtsModelConfig modelConfig = new TtsModelConfig();
+        modelConfig.setModelEnum(TtsModelEnum.SHERPA_VITS);
+        modelConfig.setModelPath("/Users/wenjie/Documents/develop/model/speech/tts/vits-zh-hf-keqing");
+        modelConfig.setModelName("keqing.onnx");
+        modelConfig.setLibPath(Paths.get("/Users/wenjie/smartjavaai_cache/sherpa-onnx-v1.12.14-osx-arm64-jni/lib"));
+        modelConfig.setDevice(device);
+        modelConfig.putCustomParam("debug", false);
+        modelConfig.putCustomParam("numThreads", 1);
+        return TtsModelFactory.getInstance().getModel(modelConfig);
+    }
+
+    /**
+     * 获取Matcha中文模型
+     * @return
+     */
+    public TtsModel getMatchaZhModel(){
+        TtsModelConfig modelConfig = new TtsModelConfig();
+        modelConfig.setModelEnum(TtsModelEnum.SHERPA_MATCHA);
+        modelConfig.setModelPath("/Users/wenjie/Documents/develop/model/speech/tts/matcha-icefall-zh-baker");
+        modelConfig.setModelName("model-steps-3.onnx");
+        modelConfig.putCustomParam("vocoder", "/Users/wenjie/Documents/develop/model/speech/tts/matcha-icefall-zh-baker/vocos-22khz-univ.onnx");
+        modelConfig.setLibPath(Paths.get("/Users/wenjie/smartjavaai_cache/sherpa-onnx-v1.12.14-osx-arm64-jni/lib"));
+        modelConfig.setDevice(device);
+        modelConfig.putCustomParam("debug", false);
+        modelConfig.putCustomParam("numThreads", 1);
+        return TtsModelFactory.getInstance().getModel(modelConfig);
+    }
+
+    /**
+     * 获取Kokoro多语言模型
+     * @return
+     */
+    public TtsModel getKokoroZhEnModel(){
+        TtsModelConfig modelConfig = new TtsModelConfig();
+        modelConfig.setModelEnum(TtsModelEnum.SHERPA_KOKORO);
+        modelConfig.setModelPath("/Users/wenjie/Documents/develop/model/speech/tts/kokoro-multi-lang-v1_0");
+        modelConfig.setModelName("model.onnx");
+        modelConfig.setLibPath(Paths.get("/Users/wenjie/smartjavaai_cache/sherpa-onnx-v1.12.14-osx-arm64-jni/lib"));
+        modelConfig.setDevice(device);
+        modelConfig.putCustomParam("debug", false);
+        modelConfig.putCustomParam("numThreads", 2);
+        return TtsModelFactory.getInstance().getModel(modelConfig);
+    }
+
+    /**
+     * 获取Matcha模型（英文）
+     * @return
+     */
+    public TtsModel getMatchaEnModel(){
+        TtsModelConfig modelConfig = new TtsModelConfig();
+        modelConfig.setModelEnum(TtsModelEnum.SHERPA_MATCHA);
+        modelConfig.setModelPath("/Users/wenjie/Documents/develop/model/speech/tts/matcha-icefall-en_US-ljspeech");
+        modelConfig.setModelName("model-steps-3.onnx");
+        modelConfig.setLibPath(Paths.get("/Users/wenjie/smartjavaai_cache/sherpa-onnx-v1.12.14-osx-arm64-jni/lib"));
+        modelConfig.putCustomParam("vocoder", "/Users/wenjie/Documents/develop/model/speech/tts/matcha-icefall-zh-baker/vocos-22khz-univ.onnx");
+        modelConfig.setDevice(device);
+        modelConfig.putCustomParam("debug", false);
+        modelConfig.putCustomParam("numThreads", 2);
+        return TtsModelFactory.getInstance().getModel(modelConfig);
+    }
+
+    /**
+     * 语音合成Vits（中文）
+     */
+    @Test
+    public void testVitsZhTts() throws IOException {
+        TtsModel ttsModel = getVitsZhModel();
+        SherpaTtsParams params = new SherpaTtsParams();
+        //语速
+        params.setSpeed(1f);
+        //说话人ID
+        params.setSpeakerId(100);
+        R<Audio> result = ttsModel.generate(zhText, params);
+        if (result.isSuccess()){
+            Audio audio = result.getData();
+            AudioUtils.saveToWav(audio.getData(),"/Users/wenjie/Downloads/tts-vits-zh.wav");
+        }else{
+            System.out.println(result.getMessage());
+        }
+    }
+
+    /**
+     * 语音合成Matcha（中文）
+     */
+    @Test
+    public void testMatchaZhTts() throws IOException {
+        TtsModel ttsModel = getMatchaZhModel();
+        SherpaTtsParams params = new SherpaTtsParams();
+        //语速
+        params.setSpeed(1.5f);
+        //说话人ID
+        params.setSpeakerId(0);
+        R<Audio> result = ttsModel.generate(zhText, params);
+        if (result.isSuccess()){
+            Audio audio = result.getData();
+            AudioUtils.saveToWav(audio.getData(),"/Users/wenjie/Downloads/tts-matcha-zh.wav");
+        }else{
+            System.out.println(result.getMessage());
+        }
+    }
+
+    /**
+     * 语音合成Kokoro（中英）
+     */
+    @Test
+    public void testKokoroZhEnTts() throws IOException {
+        TtsModel ttsModel = getKokoroZhEnModel();
+        SherpaTtsParams params = new SherpaTtsParams();
+        //语速
+        params.setSpeed(1f);
+        //说话人ID
+        params.setSpeakerId(3);
+        R<Audio> result = ttsModel.generate(znEnText, params);
+        if (result.isSuccess()){
+            Audio audio = result.getData();
+            AudioUtils.saveToWav(audio.getData(),"/Users/wenjie/Downloads/tts-kokoro-zh-en.wav");
+        }else{
+            System.out.println(result.getMessage());
+        }
+    }
+
+    /**
+     * 语音合成Matcha（英文）
+     */
+    @Test
+    public void testMatchaEnTts() throws IOException {
+        TtsModel ttsModel = getMatchaEnModel();
+        SherpaTtsParams params = new SherpaTtsParams();
+        //语速
+        params.setSpeed(1.5f);
+        //说话人ID
+        params.setSpeakerId(0);
+        R<Audio> result = ttsModel.generate(enText, params);
+        if (result.isSuccess()){
+            Audio audio = result.getData();
+            AudioUtils.saveToWav(audio.getData(),"/Users/wenjie/Downloads/tts-kitten-en.wav");
+        }else{
+            System.out.println(result.getMessage());
+        }
+    }
+
+
+}