- Hand-crafted
- Action Recognition
suvery : https://blog.csdn.net/On_theway10/article/details/88866091
- Action Localization
https://blog.csdn.net/On_theway10/article/details/89207138
[2016-CVPR] Temporal action localization in untrimmed videos via multi-stage CNNs.
[2017-ICCV] TURN TAP: Temporal unit regression network for temporal action proposals.
[2017-ICCV] R-C3D: Region convolutional 3D network for temporal activity detection.
[2017-ICCV] Temporal action detection with structured segment networks
[2017-CVPR] CDC: Convolutional-de-convolutional networks for precise temporal action localization in untrimmed videos
- Anomaly Detection
[2018-CVPR] Real-world anomaly detection in surveillance videos.
- Summarizaton/Synopsis
[2014-ECCV] Creating summaries from user videos.
[2016-TIP] Surveillance video synopsis via scaling down objects
- Video Captioning
[2015-ICCV] Sequence to Sequence–Video to Text.
- Next Frame
[2016-ICLR] Deep multi-scale video prediction beyond mean square error.
[2016-NIPS] Generating videos with scene dynamics
- Object tracking
[2016-ECCV] Fully-convolutional siamese networks for object tracking.
[2017-CVPR] ECO: Efficient Convolution Operators for tracking.
[2017-CVPR] End-to-end representation learning for correlation filter based tracking
视频理解研究进展: https://blog.csdn.net/sgfmby1994/article/details/80497537