The GTM-UVIGO System for Audiovisual Diarization 2020

Manuel Porta-Lorenzo, José Luis Alba-Castro, Laura Docío-Fernández

This paper explains in detail the Audiovisual system deployed by the Multimedia Technologies Group (GTM) of the atlanTTic research center at the University of Vigo, for the Albayzin Multimodal Diarization Challenge (MDC) organized in the Iberspeech 2020 conference. This system is characterized by the use of state of the art face and speaker verification embeddings trained with publicly available Deep Neural Networks and fine-tuned for the persons of interest. Video and audio tracks are processed separately and are finally fused to make joint decisions on the speaker diarization result.Few modifications have been made over the GTM-UVIGO system presented in the very same conference in 2018, mainly regarding the video processing part.

doi: 10.21437/IberSPEECH.2021-17

Porta-Lorenzo, M, Alba-Castro, J.L, Docío-Fernández, L (2021) The GTM-UVIGO System for Audiovisual Diarization 2020. Proc. IberSPEECH 2021, 81-85, doi: 10.21437/IberSPEECH.2021-17.