End-to-End Neural Speaker Diarization with Permutation-Free Objectives
This addresses the problem of accurately identifying and segmenting speakers in audio recordings for applications like transcription and meeting analysis, offering a simpler and more effective approach than traditional methods.
The paper tackles speaker diarization by proposing an end-to-end neural network that directly outputs results without separate modules, achieving a diarization error rate of 12.28% on simulated mixtures compared to 28.77% for a conventional system, with a 25.6% relative improvement on real data after adaptation.
In this paper, we propose a novel end-to-end neural-network-based speaker diarization method. Unlike most existing methods, our proposed method does not have separate modules for extraction and clustering of speaker representations. Instead, our model has a single neural network that directly outputs speaker diarization results. To realize such a model, we formulate the speaker diarization problem as a multi-label classification problem, and introduces a permutation-free objective function to directly minimize diarization errors without being suffered from the speaker-label permutation problem. Besides its end-to-end simplicity, the proposed method also benefits from being able to explicitly handle overlapping speech during training and inference. Because of the benefit, our model can be easily trained/adapted with real-recorded multi-speaker conversations just by feeding the corresponding multi-speaker segment labels. We evaluated the proposed method on simulated speech mixtures. The proposed method achieved diarization error rate of 12.28%, while a conventional clustering-based system produced diarization error rate of 28.77%. Furthermore, the domain adaptation with real-recorded speech provided 25.6% relative improvement on the CALLHOME dataset. Our source code is available online at https://github.com/hitachi-speech/EEND.