C api example for sense voice (#1165)

c-api-examples/CMakeLists.txt

@@ -24,8 +24,11 @@ target_link_libraries(audio-tagging-c-api sherpa-onnx-c-api)
 add_executable(add-punctuation-c-api add-punctuation-c-api.c)
 target_link_libraries(add-punctuation-c-api sherpa-onnx-c-api)
 
-add_executable(offline-stt-c-api offline-stt-c-api.c)
-target_link_libraries(offline-stt-c-api sherpa-onnx-c-api)
+add_executable(whisper-c-api whisper-c-api.c)
+target_link_libraries(whisper-c-api sherpa-onnx-c-api)
+
+add_executable(sense-voice-c-api sense-voice-c-api.c)
+target_link_libraries(sense-voice-c-api sherpa-onnx-c-api)
 
 if(SHERPA_ONNX_HAS_ALSA)
   add_subdirectory(./asr-microphone-example)

c-api-examples/sense-voice-c-api.c

@@ -0,0 +1,86 @@
+// c-api-examples/sense-voice-c-api.c
+//
+// Copyright (c)  2024  Xiaomi Corporation
+
+//
+// This file demonstrates how to use SenseVoice with sherpa-onnx's C API.
+// clang-format off
+//
+// wget https://github.com/k2-fsa/sherpa-onnx/releases/download/asr-models/sherpa-onnx-sense-voice-zh-en-ja-ko-yue-2024-07-17.tar.bz2
+// tar xvf sherpa-onnx-sense-voice-zh-en-ja-ko-yue-2024-07-17.tar.bz2
+// rm sherpa-onnx-sense-voice-zh-en-ja-ko-yue-2024-07-17.tar.bz2
+//
+// clang-format on
+
+#include <stdio.h>
+#include <stdlib.h>
+#include <string.h>
+
+#include "sherpa-onnx/c-api/c-api.h"
+
+int32_t main() {
+  // You can find more test waves from
+  // https://hf-mirror.com/spaces/k2-fsa/spoken-language-identification/tree/main/test_wavs
+  const char *wav_filename =
+      "./sherpa-onnx-sense-voice-zh-en-ja-ko-yue-2024-07-17/test_wavs/en.wav";
+  const char *model_filename =
+      "./sherpa-onnx-sense-voice-zh-en-ja-ko-yue-2024-07-17/model.int8.onnx";
+  const char *tokens_filename =
+      "./sherpa-onnx-sense-voice-zh-en-ja-ko-yue-2024-07-17/tokens.txt";
+  const char *language = "auto";
+  const char *provider = "cpu";
+  int32_t use_inverse_text_normalization = 1;
+
+  const SherpaOnnxWave *wave = SherpaOnnxReadWave(wav_filename);
+  if (wave == NULL) {
+    fprintf(stderr, "Failed to read %s\n", wav_filename);
+    return -1;
+  }
+
+  SherpaOnnxOfflineSenseVoiceModelConfig sense_voice_config;
+  memset(&sense_voice_config, 0, sizeof(sense_voice_config));
+  sense_voice_config.model = model_filename;
+  sense_voice_config.language = language;
+  sense_voice_config.use_itn = use_inverse_text_normalization;
+
+  // Offline model config
+  SherpaOnnxOfflineModelConfig offline_model_config;
+  memset(&offline_model_config, 0, sizeof(offline_model_config));
+  offline_model_config.debug = 1;
+  offline_model_config.num_threads = 1;
+  offline_model_config.provider = provider;
+  offline_model_config.tokens = tokens_filename;
+  offline_model_config.sense_voice = sense_voice_config;
+
+  // Recognizer config
+  SherpaOnnxOfflineRecognizerConfig recognizer_config;
+  memset(&recognizer_config, 0, sizeof(recognizer_config));
+  recognizer_config.decoding_method = "greedy_search";
+  recognizer_config.model_config = offline_model_config;
+
+  SherpaOnnxOfflineRecognizer *recognizer =
+      CreateOfflineRecognizer(&recognizer_config);
+
+  if (recognizer == NULL) {
+    fprintf(stderr, "Please check your config!\n");
+    SherpaOnnxFreeWave(wave);
+    return -1;
+  }
+
+  SherpaOnnxOfflineStream *stream = CreateOfflineStream(recognizer);
+
+  AcceptWaveformOffline(stream, wave->sample_rate, wave->samples,
+                        wave->num_samples);
+  DecodeOfflineStream(recognizer, stream);
+  const SherpaOnnxOfflineRecognizerResult *result =
+      GetOfflineStreamResult(stream);
+
+  fprintf(stderr, "Decoded text: %s\n", result->text);
+
+  DestroyOfflineRecognizerResult(result);
+  DestroyOfflineStream(stream);
+  DestroyOfflineRecognizer(recognizer);
+  SherpaOnnxFreeWave(wave);
+
+  return 0;
+}

c-api-examples/whisper-c-api.c

@@ -21,8 +21,6 @@
 #include "sherpa-onnx/c-api/c-api.h"
 
 int32_t main() {
-  // You can find more test waves from
-  // https://hf-mirror.com/spaces/k2-fsa/spoken-language-identification/tree/main/test_wavs
   const char *wav_filename = "./sherpa-onnx-whisper-tiny/test_wavs/0.wav";
   const char *encoder_filename = "sherpa-onnx-whisper-tiny/tiny-encoder.onnx";
   const char *decoder_filename = "sherpa-onnx-whisper-tiny/tiny-decoder.onnx";
@@ -48,31 +46,36 @@ int32_t main() {
   // Offline model config
   SherpaOnnxOfflineModelConfig offline_model_config;
   memset(&offline_model_config, 0, sizeof(offline_model_config));
-  offline_model_config.bpe_vocab = "";
   offline_model_config.debug = 1;
   offline_model_config.num_threads = 1;
   offline_model_config.provider = provider;
   offline_model_config.tokens = tokens_filename;
   offline_model_config.whisper = whisper_config;
-  offline_model_config.sense_voice =
-      (SherpaOnnxOfflineSenseVoiceModelConfig){"", "", 0};
 
   // Recognizer config
   SherpaOnnxOfflineRecognizerConfig recognizer_config;
   memset(&recognizer_config, 0, sizeof(recognizer_config));
   recognizer_config.decoding_method = "greedy_search";
-  recognizer_config.feat_config = (SherpaOnnxFeatureConfig){16000, 512};
   recognizer_config.model_config = offline_model_config;
 
   SherpaOnnxOfflineRecognizer *recognizer =
       CreateOfflineRecognizer(&recognizer_config);
 
+  if (recognizer == NULL) {
+    fprintf(stderr, "Please check your config!\n");
+
+    SherpaOnnxFreeWave(wave);
+
+    return -1;
+  }
+
   SherpaOnnxOfflineStream *stream = CreateOfflineStream(recognizer);
 
   AcceptWaveformOffline(stream, wave->sample_rate, wave->samples,
                         wave->num_samples);
   DecodeOfflineStream(recognizer, stream);
-  SherpaOnnxOfflineRecognizerResult *result = GetOfflineStreamResult(stream);
+  const SherpaOnnxOfflineRecognizerResult *result =
+      GetOfflineStreamResult(stream);
 
   fprintf(stderr, "Decoded text: %s\n", result->text);