Win-Win: Training High-Resolution Vision Transformers from Two Windows.

Vincent Leroy, Jrme Revaud, Thomas Lucas, Philippe Weinzaepfel

Browse the full ICLR paper archive.