VideoREPA: Learning Physics for Video Generation through Relational Alignment with Foundation Models

Devs

VideoREPA: Learning Physics for Video Generation through Relational Alignment with Foundation Models | Read Paper on Bytez