Rethinking Evaluation Protocols of Visual Representations Learned via Self-supervised Learning

Open in new window