Neural Speaker Diarization with Speaker-Wise Chain Rule
This solves the variable speaker count problem in audio processing for applications like transcription and meeting analysis, representing a significant but incremental improvement over existing methods.
The paper tackles the limitation of fixed speaker numbers in end-to-end neural diarization by proposing a speaker-wise conditional inference method based on the probabilistic chain rule, which allows for variable numbers of speakers and outperforms state-of-the-art methods in diarization error rate.
Speaker diarization is an essential step for processing multi-speaker audio. Although an end-to-end neural diarization (EEND) method achieved state-of-the-art performance, it is limited to a fixed number of speakers. In this paper, we solve this fixed number of speaker issue by a novel speaker-wise conditional inference method based on the probabilistic chain rule. In the proposed method, each speaker's speech activity is regarded as a single random variable, and is estimated sequentially conditioned on previously estimated other speakers' speech activities. Similar to other sequence-to-sequence models, the proposed method produces a variable number of speakers with a stop sequence condition. We evaluated the proposed method on multi-speaker audio recordings of a variable number of speakers. Experimental results show that the proposed method can correctly produce diarization results with a variable number of speakers and outperforms the state-of-the-art end-to-end speaker diarization methods in terms of diarization error rate.