Rethinking Self-supervised Correspondence Learning: A Video Frame-level Similarity Perspective.

Jiarui Xu, Xiaolong Wang

Browse the full ICCV paper archive.