v6.2

kaifcodec/user-scannerv6.2Nov 6, 2025by snakers4

AI Summary

The VAD training paradigm was reworked to improve stability and quality on various edge cases.

Key Highlights

Improved stability on out-of-distribution data.
Significant quality improvements for unusual, child, and cartoon voices.
Better handling of muted voices and low-quality phone calls.

New Features

Enhanced quality on unusual and muted voices
Improved stability on rare data

Full Release Notes

- VAD training paradigm reworked;
- Overall slight quality improvement (no metrics update);
- Higher stability on OOD / rare / strange / unique data;
- Significant quality improvements on various known edge cases:
  - Unusual voices
  - Child voices
  - Cartoon voices
  - Muted voices
  - Muted speech
  - Lower quality phone calls

## What's Changed

* Fix type hint for min_silence_at_max_speech (float -> int) by @Purfview in https://github.com/snakers4/silero-vad/pull/714
* Adamnsandle by @adamnsandle in https://github.com/snakers4/silero-vad/pull/717
* Adamnsandle by @adamnsandle in https://github.com/snakers4/silero-vad/pull/719

**Full Changelog**: https://github.com/snakers4/silero-vad/compare/v6.1...v6.2