AnotherFoxGuy/webrtc-audio-processing
1
0
Fork 0
Code Pull Requests Activity

Update common_audio

Browse Source 
Corresponds to upstream commit 524e9b043e7e86fd72353b987c9d5f6a1ebf83e1

Update notes:

 * Moved src/ to webrtc/ to easily diff against the third_party/webrtc
   in the chromium tree

 * ARM/NEON/MIPS support is not yet hooked up

 * Tests have not been copied
This commit is contained in:
Arun Raghavan
2015-10-13 12:16:16 +05:30
parent 9413986e79
commit c4fb4e38de
230 changed files with 11201 additions and 8656 deletions
Download Patch File Download Diff File Expand all files Collapse all files

115
webrtc/common_audio/vad/vad_core.h Normal file
View File

@ -0,0 +1,115 @@
/*
* Copyright (c) 2012 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.
*/
/*
* This header file includes the descriptions of the core VAD calls.
*/
#ifndef WEBRTC_COMMON_AUDIO_VAD_VAD_CORE_H_
#define WEBRTC_COMMON_AUDIO_VAD_VAD_CORE_H_
#include "webrtc/common_audio/signal_processing/include/signal_processing_library.h"
#include "webrtc/typedefs.h"
enum { kNumChannels = 6 }; // Number of frequency bands (named channels).
enum { kNumGaussians = 2 }; // Number of Gaussians per channel in the GMM.
enum { kTableSize = kNumChannels * kNumGaussians };
enum { kMinEnergy = 10 }; // Minimum energy required to trigger audio signal.
typedef struct VadInstT_
{
int vad;
int32_t downsampling_filter_states[4];
WebRtcSpl_State48khzTo8khz state_48_to_8;
int16_t noise_means[kTableSize];
int16_t speech_means[kTableSize];
int16_t noise_stds[kTableSize];
int16_t speech_stds[kTableSize];
// TODO(bjornv): Change to |frame_count|.
int32_t frame_counter;
int16_t over_hang; // Over Hang
int16_t num_of_speech;
// TODO(bjornv): Change to |age_vector|.
int16_t index_vector[16 * kNumChannels];
int16_t low_value_vector[16 * kNumChannels];
// TODO(bjornv): Change to |median|.
int16_t mean_value[kNumChannels];
int16_t upper_state[5];
int16_t lower_state[5];
int16_t hp_filter_state[4];
int16_t over_hang_max_1[3];
int16_t over_hang_max_2[3];
int16_t individual[3];
int16_t total[3];
int init_flag;
} VadInstT;
// Initializes the core VAD component. The default aggressiveness mode is
// controlled by |kDefaultMode| in vad_core.c.
//
// - self [i/o] : Instance that should be initialized
//
// returns : 0 (OK), -1 (NULL pointer in or if the default mode can't be
// set)
int WebRtcVad_InitCore(VadInstT* self);
/****************************************************************************
* WebRtcVad_set_mode_core(...)
*
* This function changes the VAD settings
*
* Input:
* - inst : VAD instance
* - mode : Aggressiveness degree
* 0 (High quality) - 3 (Highly aggressive)
*
* Output:
* - inst : Changed instance
*
* Return value : 0 - Ok
* -1 - Error
*/
int WebRtcVad_set_mode_core(VadInstT* self, int mode);
/****************************************************************************
* WebRtcVad_CalcVad48khz(...)
* WebRtcVad_CalcVad32khz(...)
* WebRtcVad_CalcVad16khz(...)
* WebRtcVad_CalcVad8khz(...)
*
* Calculate probability for active speech and make VAD decision.
*
* Input:
* - inst : Instance that should be initialized
* - speech_frame : Input speech frame
* - frame_length : Number of input samples
*
* Output:
* - inst : Updated filter states etc.
*
* Return value : VAD decision
* 0 - No active speech
* 1-6 - Active speech
*/
int WebRtcVad_CalcVad48khz(VadInstT* inst, const int16_t* speech_frame,
size_t frame_length);
int WebRtcVad_CalcVad32khz(VadInstT* inst, const int16_t* speech_frame,
size_t frame_length);
int WebRtcVad_CalcVad16khz(VadInstT* inst, const int16_t* speech_frame,
size_t frame_length);
int WebRtcVad_CalcVad8khz(VadInstT* inst, const int16_t* speech_frame,
size_t frame_length);
#endif // WEBRTC_COMMON_AUDIO_VAD_VAD_CORE_H_