Compressed Video Contrastive Learning

Existing state-of-the-art methods [Han et al. , 2020b; Tao et al. , 2020; Huo et al. , 2021] mainly focus More details can be found in Table 1. This clearly hinders large-scale video self-supervised training.