Accurate Speaker Diarization Using Channel Split

Speaker diarization using channel split offers a novel approach to address the limitations of traditional clustering methods. Instead of clustering audio segments based on acoustic features extracted from the entire audio stream, this approach leverages the spatial separation of speakers in stereo or multi-channel audio recordings.

1. Transcribe stereo audio channels

Transcribe the left and right channels of the stereo audio separately.

2. Combine the transcriptions using Subtitle Merger

Combine the transcriptions of the left and right channels.

Blogs