Android JNI support for speaker diarization (#1421)

sherpa-onnx/csrc/offline-speaker-diarization-impl.cc

@@ -23,4 +23,18 @@ OfflineSpeakerDiarizationImpl::Create(
   return nullptr;
 }
 
+#if __ANDROID_API__ >= 9
+std::unique_ptr<OfflineSpeakerDiarizationImpl>
+OfflineSpeakerDiarizationImpl::Create(
+    AAssetManager *mgr, const OfflineSpeakerDiarizationConfig &config) {
+  if (!config.segmentation.pyannote.model.empty()) {
+    return std::make_unique<OfflineSpeakerDiarizationPyannoteImpl>(mgr, config);
+  }
+
+  SHERPA_ONNX_LOGE("Please specify a speaker segmentation model.");
+
+  return nullptr;
+}
+#endif
+
 }  // namespace sherpa_onnx

sherpa-onnx/csrc/offline-speaker-diarization-impl.h

@@ -8,6 +8,11 @@
 #include <functional>
 #include <memory>
 
+#if __ANDROID_API__ >= 9
+#include "android/asset_manager.h"
+#include "android/asset_manager_jni.h"
+#endif
+
 #include "sherpa-onnx/csrc/offline-speaker-diarization.h"
 namespace sherpa_onnx {
 
@@ -16,6 +21,11 @@ class OfflineSpeakerDiarizationImpl {
   static std::unique_ptr<OfflineSpeakerDiarizationImpl> Create(
       const OfflineSpeakerDiarizationConfig &config);
 
+#if __ANDROID_API__ >= 9
+  static std::unique_ptr<OfflineSpeakerDiarizationImpl> Create(
+      AAssetManager *mgr, const OfflineSpeakerDiarizationConfig &config);
+#endif
+
   virtual ~OfflineSpeakerDiarizationImpl() = default;
 
   virtual int32_t SampleRate() const = 0;

sherpa-onnx/csrc/offline-speaker-diarization-pyannote-impl.h

@@ -10,6 +10,11 @@
 #include <utility>
 #include <vector>
 
+#if __ANDROID_API__ >= 9
+#include "android/asset_manager.h"
+#include "android/asset_manager_jni.h"
+#endif
+
 #include "Eigen/Dense"
 #include "sherpa-onnx/csrc/fast-clustering.h"
 #include "sherpa-onnx/csrc/math.h"
@@ -65,6 +70,17 @@ class OfflineSpeakerDiarizationPyannoteImpl
     Init();
   }
 
+#if __ANDROID_API__ >= 9
+  OfflineSpeakerDiarizationPyannoteImpl(
+      AAssetManager *mgr, const OfflineSpeakerDiarizationConfig &config)
+      : config_(config),
+        segmentation_model_(mgr, config_.segmentation),
+        embedding_extractor_(mgr, config_.embedding),
+        clustering_(std::make_unique<FastClustering>(config_.clustering)) {
+    Init();
+  }
+#endif
+
   int32_t SampleRate() const override {
     const auto &meta_data = segmentation_model_.GetModelMetaData();
 

sherpa-onnx/csrc/offline-speaker-diarization.cc

@@ -73,6 +73,12 @@ OfflineSpeakerDiarization::OfflineSpeakerDiarization(
     const OfflineSpeakerDiarizationConfig &config)
     : impl_(OfflineSpeakerDiarizationImpl::Create(config)) {}
 
+#if __ANDROID_API__ >= 9
+OfflineSpeakerDiarization::OfflineSpeakerDiarization(
+    AAssetManager *mgr, const OfflineSpeakerDiarizationConfig &config)
+    : impl_(OfflineSpeakerDiarizationImpl::Create(mgr, config)) {}
+#endif
+
 OfflineSpeakerDiarization::~OfflineSpeakerDiarization() = default;
 
 int32_t OfflineSpeakerDiarization::SampleRate() const {

sherpa-onnx/csrc/offline-speaker-diarization.h

@@ -9,6 +9,11 @@
 #include <memory>
 #include <string>
 
+#if __ANDROID_API__ >= 9
+#include "android/asset_manager.h"
+#include "android/asset_manager_jni.h"
+#endif
+
 #include "sherpa-onnx/csrc/fast-clustering-config.h"
 #include "sherpa-onnx/csrc/offline-speaker-diarization-result.h"
 #include "sherpa-onnx/csrc/offline-speaker-segmentation-model-config.h"
@@ -57,6 +62,11 @@ class OfflineSpeakerDiarization {
   explicit OfflineSpeakerDiarization(
       const OfflineSpeakerDiarizationConfig &config);
 
+#if __ANDROID_API__ >= 9
+  OfflineSpeakerDiarization(AAssetManager *mgr,
+                            const OfflineSpeakerDiarizationConfig &config);
+#endif
+
   ~OfflineSpeakerDiarization();
 
   // Expected sample rate of the input audio samples

sherpa-onnx/csrc/offline-speaker-segmentation-pyannote-model.cc

@@ -24,6 +24,17 @@ class OfflineSpeakerSegmentationPyannoteModel::Impl {
     Init(buf.data(), buf.size());
   }
 
+#if __ANDROID_API__ >= 9
+  Impl(AAssetManager *mgr, const OfflineSpeakerSegmentationModelConfig &config)
+      : config_(config),
+        env_(ORT_LOGGING_LEVEL_ERROR),
+        sess_opts_(GetSessionOptions(config)),
+        allocator_{} {
+    auto buf = ReadFile(mgr, config_.pyannote.model);
+    Init(buf.data(), buf.size());
+  }
+#endif
+
   const OfflineSpeakerSegmentationPyannoteModelMetaData &GetModelMetaData()
       const {
     return meta_data_;
@@ -92,6 +103,13 @@ OfflineSpeakerSegmentationPyannoteModel::
         const OfflineSpeakerSegmentationModelConfig &config)
     : impl_(std::make_unique<Impl>(config)) {}
 
+#if __ANDROID_API__ >= 9
+OfflineSpeakerSegmentationPyannoteModel::
+    OfflineSpeakerSegmentationPyannoteModel(
+        AAssetManager *mgr, const OfflineSpeakerSegmentationModelConfig &config)
+    : impl_(std::make_unique<Impl>(mgr, config)) {}
+#endif
+
 OfflineSpeakerSegmentationPyannoteModel::
     ~OfflineSpeakerSegmentationPyannoteModel() = default;
 

sherpa-onnx/csrc/offline-speaker-segmentation-pyannote-model.h

@@ -6,6 +6,11 @@
 
 #include <memory>
 
+#if __ANDROID_API__ >= 9
+#include "android/asset_manager.h"
+#include "android/asset_manager_jni.h"
+#endif
+
 #include "onnxruntime_cxx_api.h"  // NOLINT
 #include "sherpa-onnx/csrc/offline-speaker-segmentation-model-config.h"
 #include "sherpa-onnx/csrc/offline-speaker-segmentation-pyannote-model-meta-data.h"
@@ -17,6 +22,11 @@ class OfflineSpeakerSegmentationPyannoteModel {
   explicit OfflineSpeakerSegmentationPyannoteModel(
       const OfflineSpeakerSegmentationModelConfig &config);
 
+#if __ANDROID_API__ >= 9
+  OfflineSpeakerSegmentationPyannoteModel(
+      AAssetManager *mgr, const OfflineSpeakerSegmentationModelConfig &config);
+#endif
+
   ~OfflineSpeakerSegmentationPyannoteModel();
 
   const OfflineSpeakerSegmentationPyannoteModelMetaData &GetModelMetaData()

sherpa-onnx/csrc/sherpa-onnx-vad-microphone-offline-asr.cc

@@ -211,7 +211,7 @@ to download models for offline ASR.
     }
 
     while (!vad->Empty()) {
-      auto &segment = vad->Front();
+      const auto &segment = vad->Front();
       auto s = recognizer.CreateStream();
       s->AcceptWaveform(sample_rate, segment.samples.data(),
                         segment.samples.size());