Encoder-Decoder Based Attractors for End-to-End Neural Diarization
This work addresses the limitation of EEND methods in dealing with a flexible number of speakers, which is a key problem for speaker diarization tasks in applications like meeting transcription.
The paper tackles the problem of handling an unknown and flexible number of speakers in end-to-end neural diarization (EEND) by introducing an encoder-decoder-based attractor calculation module (EDA), which sequentially generates speaker-wise attractors until a stopping condition is met, and it outperforms conventional cascaded approaches in evaluations on simulated and real datasets.
This paper investigates an end-to-end neural diarization (EEND) method for an unknown number of speakers. In contrast to the conventional cascaded approach to speaker diarization, EEND methods are better in terms of speaker overlap handling. However, EEND still has a disadvantage in that it cannot deal with a flexible number of speakers. To remedy this problem, we introduce encoder-decoder-based attractor calculation module (EDA) to EEND. Once frame-wise embeddings are obtained, EDA sequentially generates speaker-wise attractors on the basis of a sequence-to-sequence method using an LSTM encoder-decoder. The attractor generation continues until a stopping condition is satisfied; thus, the number of attractors can be flexible. Diarization results are then estimated as dot products of the attractors and embeddings. The embeddings from speaker overlaps result in larger dot product values with multiple attractors; thus, this method can deal with speaker overlaps. Because the maximum number of output speakers is still limited by the training set, we also propose an iterative inference method to remove this restriction. Further, we propose a method that aligns the estimated diarization results with the results of an external speech activity detector, which enables fair comparison against cascaded approaches. Extensive evaluations on simulated and real datasets show that EEND-EDA outperforms the conventional cascaded approach.