Description
Transcribing meetings with Whisper gives loose timestamps and no speaker labels, leaving subtitles to hand-sync. WhisperX is automatic speech recognition with word-level timestamps and diarization.
Forced alignment pins every word, batched inference is fast and speakers are labeled.
Word timestamps:Precise.
Diarization:Who spoke.
Batching:Fast.
Languages:Via Whisper.
Forced alignment pins every word, batched inference is fast and speakers are labeled.
Features
Word timestamps:Precise.
Diarization:Who spoke.
Batching:Fast.
Languages:Via Whisper.
