jamosdev
/
rhubarb-lip-sync
mirror of https://github.com/DanielSWolf/rhubarb-lip-sync.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

Small VAD improvements 

* RAII
* Slightly fewer false positives
This commit is contained in:
Daniel Wolf 2016-06-24 22:35:33 +02:00
parent faa3f2b4bb
commit f275267ac7
1 changed files with 4 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

7
src/audio/voiceActivityDetection.cpp
View File

@ -6,6 +6,7 @@
#include <boost/range/adaptor/transformed.hpp> #include <boost/range/adaptor/transformed.hpp>
#include <webrtc/common_audio/vad/include/webrtc_vad.h> #include <webrtc/common_audio/vad/include/webrtc_vad.h>
#include "processing.h" #include "processing.h"
#include <gsl_util.h>
using std::vector; using std::vector;
using boost::adaptors::transformed; using boost::adaptors::transformed;
@ -19,10 +20,12 @@ BoundedTimeline<void> detectVoiceActivity(std::unique_ptr<AudioStream> audioStre
VadInst* vadHandle = WebRtcVad_Create(); VadInst* vadHandle = WebRtcVad_Create();
if (!vadHandle) throw runtime_error("Error creating WebRTC VAD handle."); if (!vadHandle) throw runtime_error("Error creating WebRTC VAD handle.");
auto freeHandle = gsl::finally([&]() { WebRtcVad_Free(vadHandle); });
int error = WebRtcVad_Init(vadHandle); int error = WebRtcVad_Init(vadHandle);
if (error) throw runtime_error("Error initializing WebRTC VAD handle."); if (error) throw runtime_error("Error initializing WebRTC VAD handle.");
const int aggressiveness = 1; // 0..3. The higher, the more is cut off. const int aggressiveness = 2; // 0..3. The higher, the more is cut off.
error = WebRtcVad_set_mode(vadHandle, aggressiveness); error = WebRtcVad_set_mode(vadHandle, aggressiveness);
if (error) throw runtime_error("Error setting WebRTC VAD aggressiveness."); if (error) throw runtime_error("Error setting WebRTC VAD aggressiveness.");
@ -39,8 +42,6 @@ BoundedTimeline<void> detectVoiceActivity(std::unique_ptr<AudioStream> audioStre
const size_t bufferCapacity = audioStream->getSampleRate() / 100; const size_t bufferCapacity = audioStream->getSampleRate() / 100;
process16bitAudioStream(*audioStream.get(), processBuffer, bufferCapacity, progressSink); process16bitAudioStream(*audioStream.get(), processBuffer, bufferCapacity, progressSink);
WebRtcVad_Free(vadHandle);
// Fill small gaps in activity // Fill small gaps in activity
const centiseconds maxGap(5); const centiseconds maxGap(5);
for (const auto& pair : getPairs(activity)) { for (const auto& pair : getPairs(activity)) {