Integrate Intelligibility with APM
- Integrates intelligibility into audio_processing.
- Allows modification of reverse stream if intelligibility enabled.
- Makes intelligibility available in audioproc_float test.
- Adds reverse stream processing to audioproc_float.
- (removed) Makes intelligibility toggleable in real time in voe_cmd_test.
- Cleans up intelligibility construction, parameters, constants and dead code.
TBR=pbos@webrtc.org
Review URL: https://codereview.webrtc.org/1234463003
Cr-Commit-Position: refs/heads/master@{#9713}
diff --git a/webrtc/modules/audio_processing/audio_processing_impl.h b/webrtc/modules/audio_processing/audio_processing_impl.h
index a44b5a8..a08f7b3 100644
--- a/webrtc/modules/audio_processing/audio_processing_impl.h
+++ b/webrtc/modules/audio_processing/audio_processing_impl.h
@@ -23,6 +23,7 @@
class AgcManagerDirect;
class AudioBuffer;
+class AudioConverter;
template<typename T>
class Beamformer;
@@ -39,6 +40,7 @@
class ProcessingComponent;
class TransientSuppressor;
class VoiceDetectionImpl;
+class IntelligibilityEnhancer;
#ifdef WEBRTC_AUDIOPROC_DEBUG_DUMP
namespace audioproc {
@@ -89,12 +91,15 @@
const StreamConfig& output_config,
float* const* dest) override;
int AnalyzeReverseStream(AudioFrame* frame) override;
+ int ProcessReverseStream(AudioFrame* frame) override;
int AnalyzeReverseStream(const float* const* data,
int samples_per_channel,
int sample_rate_hz,
ChannelLayout layout) override;
- int AnalyzeReverseStream(const float* const* data,
- const StreamConfig& reverse_config) override;
+ int ProcessReverseStream(const float* const* src,
+ const StreamConfig& reverse_input_config,
+ const StreamConfig& reverse_output_config,
+ float* const* dest) override;
int set_stream_delay_ms(int delay) override;
int stream_delay_ms() const override;
bool was_stream_delay_set() const override;
@@ -124,16 +129,23 @@
EXCLUSIVE_LOCKS_REQUIRED(crit_);
int MaybeInitializeLocked(const ProcessingConfig& config)
EXCLUSIVE_LOCKS_REQUIRED(crit_);
+ // TODO(ekm): Remove once all clients updated to new interface.
+ int AnalyzeReverseStream(const float* const* src,
+ const StreamConfig& input_config,
+ const StreamConfig& output_config);
int ProcessStreamLocked() EXCLUSIVE_LOCKS_REQUIRED(crit_);
- int AnalyzeReverseStreamLocked() EXCLUSIVE_LOCKS_REQUIRED(crit_);
+ int ProcessReverseStreamLocked() EXCLUSIVE_LOCKS_REQUIRED(crit_);
bool is_data_processed() const;
bool output_copy_needed(bool is_data_processed) const;
bool synthesis_needed(bool is_data_processed) const;
bool analysis_needed(bool is_data_processed) const;
+ bool is_rev_processed() const;
+ bool rev_conversion_needed() const;
void InitializeExperimentalAgc() EXCLUSIVE_LOCKS_REQUIRED(crit_);
void InitializeTransient() EXCLUSIVE_LOCKS_REQUIRED(crit_);
void InitializeBeamformer() EXCLUSIVE_LOCKS_REQUIRED(crit_);
+ void InitializeIntelligibility() EXCLUSIVE_LOCKS_REQUIRED(crit_);
void MaybeUpdateHistograms() EXCLUSIVE_LOCKS_REQUIRED(crit_);
EchoCancellationImpl* echo_cancellation_;
@@ -149,6 +161,7 @@
CriticalSectionWrapper* crit_;
rtc::scoped_ptr<AudioBuffer> render_audio_;
rtc::scoped_ptr<AudioBuffer> capture_audio_;
+ rtc::scoped_ptr<AudioConverter> render_converter_;
#ifdef WEBRTC_AUDIOPROC_DEBUG_DUMP
// TODO(andrew): make this more graceful. Ideally we would split this stuff
// out into a separate class with an "enabled" and "disabled" implementation.
@@ -191,6 +204,9 @@
const bool beamformer_enabled_;
rtc::scoped_ptr<Beamformer<float>> beamformer_;
const std::vector<Point> array_geometry_;
+
+ bool intelligibility_enabled_;
+ rtc::scoped_ptr<IntelligibilityEnhancer> intelligibility_enhancer_;
};
} // namespace webrtc