Cinematic audio source separation is a relatively new subtask of audio s...
We propose a dataset, AVASpeech-SMAD, to assist speech and music activit...
Vehicle re-identification (re-ID) focuses on matching images of the same...
Video-based person re-identification (Re-ID) aims at matching video sequ...
FusionGAN is a novel genre fusion framework for music generation that
in...