一作ICML 2026Spotlight· Deep Learning / Generative Models and Autoencoders
VideoFlexTok: Flexible-Length Coarse-to-Fine Video Tokenization
Andrei Atanov · Jesse Allardice · Roman Bachmann · Oğuzhan Fatih Kar · R Devon Hjelm · David Griffiths · Peter Fu · Amir Zamir · Afshin Dehghan