In the context of the Vortanz project, video-based HAR was applied to the BAST analysis.
The five main research questions that this work tackles:
- Build robust classifers for the base categories of BAST
- Build robust classifers for the evaluation categories of BAST
- Test the ability of these models to generalize to other domains
- Benefits of transfer learning in the models' performance
- Influence of background in the models' decisions
For the details checkout the presentation, or you can checkout the thesis.
MMAction2 is an open-source toolbox for video understanding based on PyTorch. It is a part of the OpenMMLab project. The master branch works with PyTorch 1.3+.
Action Recognition results on the BAST dataset
GradCAM results on SlowOnly trained without background
getting_started.md shows the basic usage of MMAction2. There are also tutorials:
- learn about configs
- finetuning models
- adding new dataset
- designing data pipeline
- adding new modules
- exporting model to onnx
- customizing runtime settings
@misc{2020mmaction2,
title={OpenMMLab's Next Generation Video Understanding Toolbox and Benchmark},
author={MMAction2 Contributors},
howpublished = {\url{https://github.com/open-mmlab/mmaction2}},
year={2020}
}@misc{mmpose2020,
title={OpenMMLab Pose Estimation Toolbox and Benchmark},
author={MMPose Contributors},
howpublished = {\url{https://github.com/open-mmlab/mmpose}},
year={2020}
}@article{mmdetection,
title = {{MMDetection}: Open MMLab Detection Toolbox and Benchmark},
author = {Chen, Kai and Wang, Jiaqi and Pang, Jiangmiao and Cao, Yuhang and
Xiong, Yu and Li, Xiaoxiao and Sun, Shuyang and Feng, Wansen and
Liu, Ziwei and Xu, Jiarui and Zhang, Zheng and Cheng, Dazhi and
Zhu, Chenchen and Cheng, Tianheng and Zhao, Qijie and Li, Buyu and
Lu, Xin and Zhu, Rui and Wu, Yue and Dai, Jifeng and Wang, Jingdong
and Shi, Jianping and Ouyang, Wanli and Loy, Chen Change and Lin, Dahua},
journal= {arXiv preprint arXiv:1906.07155},
year={2019}
}