Assessing the Robustness of Spectral Clustering for Deep Speaker Diarization

arXiv:2403.14286v1 Announce Type: cross
Abstract: Clustering speaker embeddings is crucial in speaker diarization but hasn't received as much focus as other components. Moreover, the robustness of speaker diarization across various datasets hasn't been explored when the development and evaluation data are from different domains. To bridge this gap, this study thoroughly examines spectral clustering for both same-domain and cross-domain speaker diarization. Our extensive experiments on two widely used corpora, AMI and DIHARD, reveal the performance trend of speaker diarization in …

abstract arxiv bridge clustering components cs.cv cs.lg cs.sd data datasets development diarization domains eess.as embeddings evaluation focus gap robustness speaker study type