Masked Video Distillation: Rethinking Masked Feature Modeling for Self-Supervised Video Representation Learning | Read Paper on Bytez