ViDROP: Video Dense Representation through Spatio-Temporal Sparsity.

Sepehr Sameni, Simon Jenni, Paolo Favaro

Browse the full CVPR paper archive.