Patch-level Contrastive Learning via Positional Query for Visual Pre-training.

Shaofeng Zhang, Qiang Zhou, Zhibin Wang, Fan Wang, Junchi Yan

Browse the full ICML paper archive.