1 paper
Juan Leon Alcazar, Moritz Cordes, Chen Zhao +1
Recent advances in the Active Speaker Detection (ASD) problem build upon a two-stage process: feature extraction and spatio-temporal context aggregation. In this paper, we propose…