COOKIES: By using this website you agree that we can place Google Analytics Cookies on your device for performance monitoring. |
University of Cambridge > Talks.cam > CUED Speech Group Seminars > Towards Improving End-to-End Neural Diarization
Towards Improving End-to-End Neural DiarizationAdd to your list(s) Download to your calendar using vCal
If you have a question about this talk, please contact Simon Webster McKnight. Until recently, diarization systems were formed by different submodules like voice activity detection, embedding extraction and clustering of such embeddings. However, the last quinquennial has seen many developments in diarization towards end-to-end models. These models, unlike modular ones, are trained to optimize a diarization-related loss and provide a more straightforward inference. Nevertheless, end-to-end systems still pose certain challenges. In this talk, I will comment on some of the work I did addressing some of their problems regarding synthetic training data generation and handling variable numbers of speakers. This talk is part of the CUED Speech Group Seminars series. This talk is included in these lists:
Note that ex-directory lists are not shown. |
Other listsType the title of a new list here Persian Society talks Wall Street meets Lincoln's Inn!Other talksFree Programme/Sightseeing Nanofluidics and Fluid Transport on Excitable Surfaces Exact phylodynamics via structured Markov genealogy processes On a skew stable Lévy process (Lecture 3) |