OpenGVLab / VideoMAEv2

[CVPR 2023] VideoMAE V2: Scaling Video Masked Autoencoders with Dual Masking
https://arxiv.org/abs/2303.16727
MIT License
524 stars 63 forks source link

could you please provide me the weight of VideoMAEv2 pre-trained on Kinetics-400? #3

Closed Value-Jack closed 1 year ago

Value-Jack commented 1 year ago

I want to use the the weight to extract the features of the thumos14. I'll appreciate it very much if you could share me the weights or the features of the thumos14 extracted by Kinetics-400 !

congee524 commented 1 year ago

Hi, the weights are available at ModelZoo

fmthoker commented 1 year ago

@wanglimin The shared models do not contain any model with Kinetics-400 pretraining. Can you share a ViT-B model pre-trained on Kinetics-400 with dual masking? I am doing a comparison of Video SSL models and want to include Videomae and Videomaev2 with the same backbone and pretraining dataset.