audio-speaker-tools - Speaker Separation and Audio Processing Tools
Separate speakers, compare voices, and process audio files using Demucs, pyannote, and Resemblyzer
Tags
Updated: 2026-03-24Capabilities
Typical Inputs
Typical Outputs
What this skill does
- separate speakers from audio
- compare voice samples
- extract audio segments
- isolate vocals
- validate speaker diarization
- prepare voice samples
Inputs
- audio files
- HuggingFace token
- speaker count range
Outputs
- separated speaker audio files
- voice similarity scores
- audio segments
- diarization metadata files
Requirements
- Python 3.9+
- ffmpeg
- PyTorch MPS support
- HuggingFace token