Official Open Source code for "Masked Autoencoders As Spatiotemporal Learners"

[CVPR2023] Masked Video Distillation: Rethinking Masked Feature Modeling for Self-supervised Video Representation Learning (https://arxiv.org/abs/2212.04500)

[CVPR 2023] VideoMAE V2: Scaling Video Masked Autoencoders with Dual Masking

Hiera: A fast, powerful, and simple hierarchical vision transformer.