Online Neural Diarization of Unlimited Numbers of Speakers

Horiguchi, Shota; Watanabe, Soichi; Garcia, Paola; Takashima, Yuki; Kawaguchi, Yohei

doi:10.48550/arxiv.2206.02432

Cited by 1 publication

(1 citation statement)

References 56 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Early works on online SD adopted a universal background model or an i-vector with log-likelihood or cosine similarity-based clustering [13,14]. With advances in deep learning, more recent works rely upon either speaker embeddings extracted from a deep neural network (DNN) or entirely compose an online end-to-end neural diarisation (EEND) models [18,19,21]. Xue et al [19] introduced EEND-based online diarisation systems using speaker-tracing buffer (STB).…”

Section: Introductionmentioning

confidence: 99%

Absolute decision corrupts absolutely: conservative online speaker diarisation

Kwon¹,

Heo²,

Lee³

et al. 2022

Preprint

View full text Add to dashboard Cite

Our focus lies in developing an online speaker diarisation framework which demonstrates robust performance across diverse domains. In online speaker diarisation, outputs generated in real-time are irreversible, and a few misjudgements in the early phase of an input session can lead to catastrophic results. We hypothesise that cautiously increasing the number of estimated speakers is of paramount importance among many other factors. Thus, our proposed framework includes decreasing the number of speakers by one when the system judges that an increase in the past was faulty. We also adopt dual buffers, checkpoints and centroids, where checkpoints are combined with silhouette coefficients to estimate the number of speakers and centroids represent speakers. Again, we believe that more than one centroid can be generated from one speaker. Thus we design a clustering-based label matching technique to assign labels in realtime. The resulting system is lightweight yet surprisingly effective. The system demonstrates state-of-the-art performance on DIHARD II and III datasets, where it is also competitive in AMI and VoxConverse test sets.

show abstract