Vad patch #1369
Open
Vad patch #1369
+306
−119
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
|
|
OK, I'll sort it out. The another LFR that affects the recognition result can be processed separately. |
csukuangfj
requested changes
Oct 29, 2024
| @@ -102,12 +101,12 @@ bool SileroVadModelConfig::Validate() const { | |||
| std::string SileroVadModelConfig::ToString() const { | |||
| std::ostringstream os; | |||
|
|
|||
| os << "SileroVadModelConfig("; | |||
| os << "SilerVadModelConfig("; | |||
There was a problem hiding this comment.
please don't change it.
| @@ -31,8 +31,7 @@ void SileroVadModelConfig::Register(ParseOptions *po) { | |||
| po->Register( | |||
| "silero-vad-max-speech-duration", &max_speech_duration, | |||
| "In seconds. If a speech segment is longer than this value, then we " | |||
| "increase the threshold to 0.9. After finishing detecting the segment, " | |||
| "the threshold value is reset to its original value."); | |||
| "cut a segment."); | |||
There was a problem hiding this comment.
please don't remove it.
| // the threshold to 0.9. After finishing detecting the segment, | ||
| // the threshold value is reset to its original value. | ||
| float max_speech_duration = 20; // in seconds | ||
| float max_speech_duration = 20; // in seconds |
There was a problem hiding this comment.
please don't remove the comments.
Comment on lines
35
to
+42
| min_silence_samples_ = | ||
| sample_rate_ * config_.silero_vad.min_silence_duration; | ||
| (int32_t)(sample_rate_ * config_.silero_vad.min_silence_duration); | ||
|
|
||
| min_speech_samples_ = sample_rate_ * config_.silero_vad.min_speech_duration; | ||
| min_speech_samples_ = | ||
| (int32_t)(sample_rate_ * config_.silero_vad.min_speech_duration); | ||
|
|
||
| max_speech_samples_ = | ||
| (int32_t)(sample_rate_ * config_.silero_vad.max_speech_duration); |
There was a problem hiding this comment.
Is there a reason to make such changes?
|
By the way, could you describe the issue this PR tries to fix? |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Optimized the sharding logic, the SpeechSegment time period is controlled in the voice-activity-detector module, and the export of the last flush is more concise