Use VAD to get a better speech power estimation in the IntelligibilityEnhancer

webrtc/modules/audio_processing/intelligibility/intelligibility_enhancer.h

Index: webrtc/modules/audio_processing/intelligibility/intelligibility_enhancer.h
diff --git a/webrtc/modules/audio_processing/intelligibility/intelligibility_enhancer.h b/webrtc/modules/audio_processing/intelligibility/intelligibility_enhancer.h
index fade1449ccdb3886af3dae5154fef3ee6f176bac..be20e11d50bb36b8d1e547d7c9638944ee0711fb 100644
--- a/webrtc/modules/audio_processing/intelligibility/intelligibility_enhancer.h
+++ b/webrtc/modules/audio_processing/intelligibility/intelligibility_enhancer.h
@@ -8,10 +8,6 @@
  *  be found in the AUTHORS file in the root of the source tree.
  */
 
-//
-//  Specifies core class for intelligbility enhancement.
-//
-
 #ifndef WEBRTC_MODULES_AUDIO_PROCESSING_INTELLIGIBILITY_INTELLIGIBILITY_ENHANCER_H_
 #define WEBRTC_MODULES_AUDIO_PROCESSING_INTELLIGIBILITY_INTELLIGIBILITY_ENHANCER_H_
 
@@ -22,43 +18,18 @@
 #include "webrtc/common_audio/lapped_transform.h"
 #include "webrtc/common_audio/channel_buffer.h"
 #include "webrtc/modules/audio_processing/intelligibility/intelligibility_utils.h"
+#include "webrtc/modules/audio_processing/vad/voice_activity_detector.h"
 
 namespace webrtc {
 
 // Speech intelligibility enhancement module. Reads render and capture
 // audio streams and modifies the render stream with a set of gains per
 // frequency bin to enhance speech against the noise background.
-// Note: assumes speech and noise streams are already separated.
+// Details of the model and algorithm can be found in the original paper:
+// http://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=6882788
 class IntelligibilityEnhancer {
  public:
-  struct Config {
-    // |var_*| are parameters for the VarianceArray constructor for the
-    // clear speech stream.
-    // TODO(bercic): the |var_*|, |*_rate| and |gain_limit| parameters should
-    // probably go away once fine tuning is done.
-    Config()
-        : sample_rate_hz(16000),
-          num_capture_channels(1),
-          num_render_channels(1),
-          var_type(intelligibility::VarianceArray::kStepDecaying),
-          var_decay_rate(0.9f),
-          var_window_size(10),
-          analysis_rate(800),
-          gain_change_limit(0.1f),
-          rho(0.02f) {}
-    int sample_rate_hz;
-    size_t num_capture_channels;
-    size_t num_render_channels;
-    intelligibility::VarianceArray::StepType var_type;
-    float var_decay_rate;
-    size_t var_window_size;
-    int analysis_rate;
-    float gain_change_limit;
-    float rho;
-  };
-
-  explicit IntelligibilityEnhancer(const Config& config);
-  IntelligibilityEnhancer();  // Initialize with default config.
+  IntelligibilityEnhancer(int sample_rate_hz, size_t num_render_channels);
 
   // Sets the capture noise magnitude spectrum estimate.
   void SetCaptureNoiseEstimate(std::vector<float> noise);
@@ -90,14 +61,11 @@ class IntelligibilityEnhancer {
   FRIEND_TEST_ALL_PREFIXES(IntelligibilityEnhancerTest, TestErbCreation);
   FRIEND_TEST_ALL_PREFIXES(IntelligibilityEnhancerTest, TestSolveForGains);
 
-  // Updates variance computation and analysis with |in_block_|,
+  // Updates power computation and analysis with |in_block_|,
   // and writes modified speech to |out_block|.
   void ProcessClearBlock(const std::complex<float>* in_block,
                          std::complex<float>* out_block);
 
-  // Computes and sets modified gains.
-  void AnalyzeClearBlock(float power_target);
-
   // Bisection search for optimal |lambda|.
   void SolveForLambda(float power_target, float power_bot, float power_top);
 
@@ -114,29 +82,24 @@ class IntelligibilityEnhancer {
   // Negative gains are set to 0. Stores the results in |sols|.
   void SolveForGainsGivenLambda(float lambda, size_t start_freq, float* sols);
 
+  // Returns true if the audio is speech.
+  bool IsSpeech(const float* audio);
+
   const size_t freqs_;         // Num frequencies in frequency domain.
-  const size_t window_size_;   // Window size in samples; also the block size.
   const size_t chunk_length_;  // Chunk size in samples.
   const size_t bank_size_;     // Num ERB filters.
   const int sample_rate_hz_;
-  const int erb_resolution_;
-  const size_t num_capture_channels_;
   const size_t num_render_channels_;
-  const int analysis_rate_;    // Num blocks before gains recalculated.
-
-  const bool active_;          // Whether render gains are being updated.
-                               // TODO(ekm): Add logic for updating |active_|.
 
-  intelligibility::VarianceArray clear_variance_;
-  std::vector<float> noise_power_;
-  rtc::scoped_ptr<float[]> filtered_clear_var_;
-  rtc::scoped_ptr<float[]> filtered_noise_var_;
+  intelligibility::PowerEstimator clear_power_estimator_;
+  rtc::scoped_ptr<intelligibility::PowerEstimator> noise_power_estimator_;
+  rtc::scoped_ptr<float[]> filtered_clear_pow_;
+  rtc::scoped_ptr<float[]> filtered_noise_pow_;
   rtc::scoped_ptr<float[]> center_freqs_;
   std::vector<std::vector<float>> capture_filter_bank_;
   std::vector<std::vector<float>> render_filter_bank_;
   size_t start_freq_;
-  rtc::scoped_ptr<float[]> rho_;  // Production and interpretation SNR.
-                                  // for each ERB band.
+
   rtc::scoped_ptr<float[]> gains_eq_;  // Pre-filter modified gains.
   intelligibility::GainApplier gain_applier_;
 
@@ -144,11 +107,13 @@ class IntelligibilityEnhancer {
   // the original input array with modifications.
   ChannelBuffer<float> temp_render_out_buffer_;
 
-  rtc::scoped_ptr<float[]> kbd_window_;
   TransformCallback render_callback_;
   rtc::scoped_ptr<LappedTransform> render_mangler_;
-  int block_count_;
-  int analysis_step_;
+
+  VoiceActivityDetector vad_;
+  std::vector<int16_t> audio_s16_;
+  size_t chunks_since_voice_;
+  bool is_speech_;
 };
 
 }  // namespace webrtc