Official PyTorch Implementation of VideoMAE (NeurIPS 2022 Spotlight).

VideoMAE: Masked Autoencoders are Data-Efficient Learners for Self-Supervised Video Pre-Training
Zhan Tong, Yibing Song, Jue Wang, Limin Wang
Nanjing University, Tencent AI Lab

📰 News

[2023.4.18] 🎈Everyone can download Kinetics-400, which is used in VideoMAE, from this link.
[2023.4.18] Code and pre-trained models of VideoMAE V2 have been released! Check and enjoy this repo!
[2023.4.17] We propose EVAD, an end-to-end Video Action Detection framework.
[2023.2.28] Our VideoMAE V2 is accepted by CVPR 2023! 🎉
[2023.1.16] Code and pre-trained models for Action Detection in VideoMAE are available!
[2022.12.27] 🎈Everyone can download extracted VideoMAE features of THUMOS, ActivityNet, HACS and FineAction from InternVideo.
[2022.11.20] 👀 VideoMAE is integrated into and , supported by @Sayak Paul.
[2022.10.25] 👀 VideoMAE is integrated into MMAction2, the results on Kinetics-400 can be reproduced successfully.
[2022.10.20] The pre-trained models and scripts of ViT-S and ViT-H are available!
[2022.10.19] The pre-trained models and scripts on UCF101 are available!
[2022.9.15] VideoMAE is accepted by NeurIPS 2022 as a spotlight presentation! 🎉
[2022.8.8] 👀 VideoMAE is integrated into official 🤗HuggingFace Transformers now!
[2022.7.7] We have updated new results on downstream AVA 2.2 benchmark. Please refer to our paper for details.
[2022.4.24] Code and pre-trained models are available now!
[2022.3.24] Code and pre-trained models will be released here. Welcome to watch this repository for the latest updates.

✨ Highlights

🔥 Masked Video Modeling for Video Pre-Training

VideoMAE performs the task of masked video modeling for video pre-training. We propose the extremely high masking ratio (90%-95%) and tube masking strategy to create a challenging task for self-supervised video pre-training.

⚡️ A Simple, Efficient and Strong Baseline in SSVP

VideoMAE uses the simple masked autoencoder and plain ViT backbone to perform video self-supervised learning. Due to the extremely high masking ratio, the pre-training time of VideoMAE is much shorter than contrastive learning methods (3.2x speedup). VideoMAE can serve as a simple but strong baseline for future research in self-supervised video pre-training.

😮 High performance, but NO extra data required

VideoMAE works well for video datasets of different scales and can achieve 87.4% on Kinects-400, 75.4% on Something-Something V2, 91.3% on UCF101, and 62.6% on HMDB51. To our best knowledge, VideoMAE is the first to achieve the state-of-the-art performance on these four popular benchmarks with the vanilla ViT backbones while doesn't need any extra data or pre-trained models.

🚀 Main Results

✨ Something-Something V2

Method	Extra Data	Backbone	Resolution	#Frames x Clips x Crops	Top-1	Top-5
VideoMAE	no	ViT-S	224x224	16x2x3	66.8	90.3
VideoMAE	no	ViT-B	224x224	16x2x3	70.8	92.4
VideoMAE	no	ViT-L	224x224	16x2x3	74.3	94.6
VideoMAE	no	ViT-L	224x224	32x1x3	75.4	95.2

✨ Kinetics-400

Method	Extra Data	Backbone	Resolution	#Frames x Clips x Crops	Top-1	Top-5
VideoMAE	no	ViT-S	224x224	16x5x3	79.0	93.8
VideoMAE	no	ViT-B	224x224	16x5x3	81.5	95.1
VideoMAE	no	ViT-L	224x224	16x5x3	85.2	96.8
VideoMAE	no	ViT-H	224x224	16x5x3	86.6	97.1
VideoMAE	no	ViT-L	320x320	32x4x3	86.1	97.3
VideoMAE	no	ViT-H	320x320	32x4x3	87.4	97.6

✨ AVA 2.2

Please check the code and checkpoints in VideoMAE-Action-Detection.

Method	Extra Data	Extra Label	Backbone	#Frame x Sample Rate	mAP
VideoMAE	Kinetics-400	✗	ViT-S	16x4	22.5
VideoMAE	Kinetics-400	✓	ViT-S	16x4	28.4
VideoMAE	Kinetics-400	✗	ViT-B	16x4	26.7
VideoMAE	Kinetics-400	✓	ViT-B	16x4	31.8
VideoMAE	Kinetics-400	✗	ViT-L	16x4	34.3
VideoMAE	Kinetics-400	✓	ViT-L	16x4	37.0
VideoMAE	Kinetics-400	✗	ViT-H	16x4	36.5
VideoMAE	Kinetics-400	✓	ViT-H	16x4	39.5
VideoMAE	Kinetics-700	✗	ViT-L	16x4	36.1
VideoMAE	Kinetics-700	✓	ViT-L	16x4	39.3

✨ UCF101 & HMDB51

Method	Extra Data	Backbone	UCF101	HMDB51
VideoMAE	no	ViT-B	91.3	62.6
VideoMAE	Kinetics-400	ViT-B	96.1	73.3

🔨 Installation

Please follow the instructions in INSTALL.md.

➡️ Data Preparation

Please follow the instructions in DATASET.md for data preparation.

🔄 Pre-training

The pre-training instruction is in PRETRAIN.md.

⤴️ Fine-tuning with pre-trained models

The fine-tuning instruction is in FINETUNE.md.

📍Model Zoo

We provide pre-trained and fine-tuned models in MODEL_ZOO.md.

👀 Visualization

We provide the script for visualization in vis.sh. Colab notebook for better visualization is coming soon.

☎️ Contact

Zhan Tong: tongzhan@smail.nju.edu.cn

👍 Acknowledgements

Thanks to Ziteng Gao, Lei Chen, Chongjian Ge, and Zhiyu Zhao for their kind support.
This project is built upon MAE-pytorch and BEiT. Thanks to the contributors of these great codebases.

🔒 License

The majority of this project is released under the CC-BY-NC 4.0 license as found in the LICENSE file. Portions of the project are available under separate license terms: SlowFast and pytorch-image-models are licensed under the Apache 2.0 license. BEiT is licensed under the MIT license.

✏️ Citation

If you think this project is helpful, please feel free to leave a star⭐️ and cite our paper:

@inproceedings{tong2022videomae,
  title={Video{MAE}: Masked Autoencoders are Data-Efficient Learners for Self-Supervised Video Pre-Training},
  author={Zhan Tong and Yibing Song and Jue Wang and Limin Wang},
  booktitle={Advances in Neural Information Processing Systems},
  year={2022}
}

@article{videomae,
  title={VideoMAE: Masked Autoencoders are Data-Efficient Learners for Self-Supervised Video Pre-Training},
  author={Tong, Zhan and Song, Yibing and Wang, Jue and Wang, Limin},
  journal={arXiv preprint arXiv:2203.12602},
  year={2022}
}