Add new variance update option and unittests for intelligibility

webrtc/modules/audio_processing/intelligibility/intelligibility_enhancer.cc

 /*
  *  Copyright (c) 2014 The WebRTC project authors. All Rights Reserved.
  *
  *  Use of this source code is governed by a BSD-style license
  *  that can be found in the LICENSE file in the root of the source
  *  tree. An additional intellectual property rights grant can be found
  *  in the file PATENTS.  All contributing project authors may
  *  be found in the AUTHORS file in the root of the source tree.
  */
 
@@ skipping 23 matching lines @@
 namespace webrtc {
 
 const int IntelligibilityEnhancer::kErbResolution = 2;
 const int IntelligibilityEnhancer::kWindowSizeMs = 2;
 const int IntelligibilityEnhancer::kChunkSizeMs = 10;  // Size provided by APM.
 const int IntelligibilityEnhancer::kAnalyzeRate = 800;
 const int IntelligibilityEnhancer::kVarianceRate = 2;
 const float IntelligibilityEnhancer::kClipFreq = 200.0f;
 const float IntelligibilityEnhancer::kConfigRho = 0.02f;
 const float IntelligibilityEnhancer::kKbdAlpha = 1.5f;
 

[hlundin-webrtc, 2015/06/30 14:00:52]

Delete empty line.

[ekm, 2015/07/01 23:48:25]

Done.

+const float IntelligibilityEnhancer::kLambdaBot = -1.0;

[hlundin-webrtc, 2015/06/30 14:00:52]

Are all of these constants only used locally in this .cc file? If so, you can
move them out of the class and simply declare them in an anonymous namespace in
this file. E.g.,

namespace {
static const int kErbResolution = 2;
...
}  // namespace

[ekm, 2015/07/01 23:48:25]

Done.

+const float IntelligibilityEnhancer::kLambdaTop = -10e-18f;
+
 // To disable gain update smoothing, set gain limit to be VERY high.
 // TODO(ekmeyerson): Add option to disable gain smoothing altogether
 // to avoid the extra computation.
 const float IntelligibilityEnhancer::kGainChangeLimit = 0.0125f;
 
 using VarianceType = intelligibility::VarianceArray::StepType;
 
 IntelligibilityEnhancer::TransformCallback::TransformCallback(
     IntelligibilityEnhancer* parent,
     IntelligibilityEnhancer::AudioSource source)
@@ skipping 177 matching lines @@
   }
 
   /* efidata(n,:) = sqrt(b(n)) * fidata(n,:) */
   gain_applier_.Apply(in_block, out_block);
 }
 
 void IntelligibilityEnhancer::AnalyzeClearBlock(float power_target) {
   FilterVariance(clear_variance_.variance(), filtered_clear_var_.get());
   FilterVariance(noise_variance_.variance(), filtered_noise_var_.get());
 
-  // Bisection search for optimal |lambda|
-
-  float lambda_bot = -1.0f, lambda_top = -10e-18f, lambda;
-  float power_bot, power_top, power;
-  SolveForGainsGivenLambda(lambda_top, start_freq_, gains_eq_.get());
+  float power_bot, power_top;

[hlundin-webrtc, 2015/06/30 14:00:51]

Premature declaration of variables. Declare them when they are used:
float power_top = DotProduct(...);

[ekm, 2015/07/01 23:48:25]

Done.

+  SolveForGainsGivenLambda(kLambdaTop, start_freq_, gains_eq_.get());
   power_top =
       DotProduct(gains_eq_.get(), filtered_clear_var_.get(), bank_size_);
-  SolveForGainsGivenLambda(lambda_bot, start_freq_, gains_eq_.get());
+  SolveForGainsGivenLambda(kLambdaBot, start_freq_, gains_eq_.get());
   power_bot =
       DotProduct(gains_eq_.get(), filtered_clear_var_.get(), bank_size_);
-  DCHECK(power_target >= power_bot && power_target <= power_top);
+  if(power_target >= power_bot && power_target <= power_top) {
+    SolveForLambda(power_target, power_bot, power_top);
+    UpdateErbGains();
+  }  // Else experiencing variance underflow, so do nothing.
+}
 
+void IntelligibilityEnhancer::SolveForLambda(float power_target,
+                                             float power_bot,
+                                             float power_top) {
+  float lambda_bot = kLambdaBot;
+  float lambda_top = kLambdaTop;
+  float lambda, power;

[hlundin-webrtc, 2015/06/30 14:00:52]

Skip declaration of lambda and power here and declare them when used inside the
loop.

[ekm, 2015/07/01 23:48:25]

Done.

   float power_ratio = 2.0f;  // Ratio of achieved power to target power.
   const float kConvergeThresh = 0.001f;  // TODO(ekmeyerson): Find best values
   const int kMaxIters = 100;             // for these, based on experiments.
   int iters = 0;
   while (fabs(power_ratio - 1.0f) > kConvergeThresh && iters <= kMaxIters) {
     lambda = lambda_bot + (lambda_top - lambda_bot) / 2.0f;
     SolveForGainsGivenLambda(lambda, start_freq_, gains_eq_.get());
     power = DotProduct(gains_eq_.get(), filtered_clear_var_.get(), bank_size_);
     if (power < power_target) {
       lambda_bot = lambda;
     } else {
       lambda_top = lambda;
     }
     power_ratio = fabs(power / power_target);
     ++iters;
   }
+}
 
+void IntelligibilityEnhancer::UpdateErbGains() {
   // (ERB gain) = filterbank' * (freq gain)
   float* gains = gain_applier_.target();
   for (int i = 0; i < freqs_; ++i) {
     gains[i] = 0.0f;
     for (int j = 0; j < bank_size_; ++j) {
       gains[i] = fmaf(filter_bank_[j][i], gains_eq_[j], gains[i]);
     }
   }
 }
 
@@ skipping 116 matching lines @@
                                           int length) {
   float ret = 0.0f;
 
   for (int i = 0; i < length; ++i) {
     ret = fmaf(a[i], b[i], ret);
   }
   return ret;
 }
 
 }  // namespace webrtc