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

webrtc/modules/audio_processing/intelligibility/test/intelligibility_proc.cc

Index: webrtc/modules/audio_processing/intelligibility/test/intelligibility_proc.cc
diff --git a/webrtc/modules/audio_processing/intelligibility/test/intelligibility_proc.cc b/webrtc/modules/audio_processing/intelligibility/test/intelligibility_proc.cc
index e02e64e709e1e01c0c191fea24af2a402592572d..185f17c4cb76cdf1e3984306cbe5517a5d9b7ab2 100644
--- a/webrtc/modules/audio_processing/intelligibility/test/intelligibility_proc.cc
+++ b/webrtc/modules/audio_processing/intelligibility/test/intelligibility_proc.cc
@@ -24,6 +24,7 @@
 #include "testing/gtest/include/gtest/gtest.h"
 #include "webrtc/base/checks.h"
 #include "webrtc/base/criticalsection.h"
+#include "webrtc/common_audio/include/audio_util.h"
 #include "webrtc/common_audio/real_fourier.h"
 #include "webrtc/common_audio/wav_file.h"
 #include "webrtc/modules/audio_processing/audio_buffer.h"
@@ -35,36 +36,10 @@
 #include "webrtc/test/testsupport/fileutils.h"
 
 using std::complex;
-using webrtc::intelligibility::VarianceArray;
 
 namespace webrtc {
 namespace {
 
-bool ValidateClearWindow(const char* flagname, int32_t value) {
-  return value > 0;
-}
-
-DEFINE_int32(clear_type,
-             webrtc::intelligibility::VarianceArray::kStepDecaying,
-             "Variance algorithm for clear data.");
-DEFINE_double(clear_alpha, 0.9, "Variance decay factor for clear data.");
-DEFINE_int32(clear_window,
-             475,
-             "Window size for windowed variance for clear data.");
-const bool clear_window_dummy =
-    google::RegisterFlagValidator(&FLAGS_clear_window, &ValidateClearWindow);
-DEFINE_int32(sample_rate,
-             16000,
-             "Audio sample rate used in the input and output files.");
-DEFINE_int32(ana_rate,
-             800,
-             "Analysis rate; gains recalculated every N blocks.");
-DEFINE_int32(
-    var_rate,
-    2,
-    "Variance clear rate; history is forgotten every N gain recalculations.");
-DEFINE_double(gain_limit, 1000.0, "Maximum gain change in one block.");
-
 DEFINE_string(clear_file, "speech.wav", "Input file with clear speech.");
 DEFINE_string(noise_file, "noise.wav", "Input file with noise data.");
 DEFINE_string(out_file,
@@ -72,64 +47,45 @@ DEFINE_string(out_file,
               "Enhanced output. Use '-' to "
               "play through aplay immediately.");
 
-const size_t kNumChannels = 1;
-
 // void function for gtest
 void void_main(int argc, char* argv[]) {
   google::SetUsageMessage(
-      "\n\nVariance algorithm types are:\n"
-      "  0 - infinite/normal,\n"
-      "  1 - exponentially decaying,\n"
-      "  2 - rolling window.\n"
-      "\nInput files must be little-endian 16-bit signed raw PCM.\n");
+      "\n\nInput files must be little-endian 16-bit signed raw PCM.\n");
   google::ParseCommandLineFlags(&argc, &argv, true);
 
-  size_t samples;        // Number of samples in input PCM file
-  size_t fragment_size;  // Number of samples to process at a time
-                         // to simulate APM stream processing
-
   // Load settings and wav input.
-
-  fragment_size = FLAGS_sample_rate / 100;  // Mirror real time APM chunk size.
-                                            // Duplicates chunk_length_ in
-                                            // IntelligibilityEnhancer.
-
   struct stat in_stat, noise_stat;
   ASSERT_EQ(stat(FLAGS_clear_file.c_str(), &in_stat), 0)
       << "Empty speech file.";
   ASSERT_EQ(stat(FLAGS_noise_file.c_str(), &noise_stat), 0)
       << "Empty noise file.";
 
-  samples = std::min(in_stat.st_size, noise_stat.st_size) / 2;
+  const size_t samples = std::min(in_stat.st_size, noise_stat.st_size) / 2;
 
   WavReader in_file(FLAGS_clear_file);
   std::vector<float> in_fpcm(samples);
   in_file.ReadSamples(samples, &in_fpcm[0]);
+  FloatS16ToFloat(&in_fpcm[0], samples, &in_fpcm[0]);
 
   WavReader noise_file(FLAGS_noise_file);
   std::vector<float> noise_fpcm(samples);
   noise_file.ReadSamples(samples, &noise_fpcm[0]);
+  FloatS16ToFloat(&noise_fpcm[0], samples, &noise_fpcm[0]);
 
   // Run intelligibility enhancement.
-  IntelligibilityEnhancer::Config config;
-  config.sample_rate_hz = FLAGS_sample_rate;
-  config.var_type = static_cast<VarianceArray::StepType>(FLAGS_clear_type);
-  config.var_decay_rate = static_cast<float>(FLAGS_clear_alpha);
-  config.var_window_size = static_cast<size_t>(FLAGS_clear_window);
-  config.analysis_rate = FLAGS_ana_rate;
-  config.gain_change_limit = FLAGS_gain_limit;
-  IntelligibilityEnhancer enh(config);
+  IntelligibilityEnhancer enh(in_file.sample_rate(), in_file.num_channels());
   rtc::CriticalSection crit;
   NoiseSuppressionImpl ns(&crit);
-  ns.Initialize(kNumChannels, FLAGS_sample_rate);
+  ns.Initialize(noise_file.num_channels(), noise_file.sample_rate());
   ns.Enable(true);
 
-  AudioBuffer capture_audio(fragment_size,
-                            kNumChannels,
-                            fragment_size,
-                            kNumChannels,
+  // Mirror real time APM chunk size. Duplicates chunk_length_ in
+  // IntelligibilityEnhancer.
+  size_t fragment_size = in_file.sample_rate() / 100;
+  AudioBuffer capture_audio(fragment_size, noise_file.num_channels(),
+                            fragment_size, noise_file.num_channels(),
                             fragment_size);
-  StreamConfig stream_config(FLAGS_sample_rate, kNumChannels);
+  StreamConfig stream_config(in_file.sample_rate(), noise_file.num_channels());
 
   // Slice the input into smaller chunks, as the APM would do, and feed them
   // through the enhancer.
@@ -141,22 +97,27 @@ void void_main(int argc, char* argv[]) {
     ns.AnalyzeCaptureAudio(&capture_audio);
     ns.ProcessCaptureAudio(&capture_audio);
     enh.SetCaptureNoiseEstimate(ns.NoiseEstimate());
-    enh.ProcessRenderAudio(&clear_cursor, FLAGS_sample_rate, kNumChannels);
+    enh.ProcessRenderAudio(&clear_cursor, in_file.sample_rate(),
+                           in_file.num_channels());
     clear_cursor += fragment_size;
     noise_cursor += fragment_size;
   }
 
+  FloatToFloatS16(&in_fpcm[0], samples, &in_fpcm[0]);
+
   if (FLAGS_out_file.compare("-") == 0) {
     const std::string temp_out_filename =
         test::TempFilename(test::WorkingDir(), "temp_wav_file");
     {
-      WavWriter out_file(temp_out_filename, FLAGS_sample_rate, kNumChannels);
+      WavWriter out_file(temp_out_filename, in_file.sample_rate(),
+                         in_file.num_channels());
       out_file.WriteSamples(&in_fpcm[0], samples);
     }
     system(("aplay " + temp_out_filename).c_str());
     system(("rm " + temp_out_filename).c_str());
   } else {
-    WavWriter out_file(FLAGS_out_file, FLAGS_sample_rate, kNumChannels);
+    WavWriter out_file(FLAGS_out_file, in_file.sample_rate(),
+                       in_file.num_channels());
     out_file.WriteSamples(&in_fpcm[0], samples);
   }
 }