Multi-Speaker Separation

Multi-speaker separation splits audio containing multiple, often overlapping, talkers into separate tracks, one per voice. It is essential for transcription accuracy, dubbing, accessibility, and meeting or call-center analytics, wherever individual voices must be edited or attributed. It is closely related to, but distinct from, speaker diarization, which labels who spoke when without necessarily isolating each voice.