Optical Flow Guided Feature: A Fast and Robust Motion Representation for Video Action Recognition

@article{Sun2018OpticalFG,
  title={Optical Flow Guided Feature: A Fast and Robust Motion Representation for Video Action Recognition},
  author={S. Sun and Zhanghui Kuang and Wanli Ouyang and Lu Sheng and Wayne Zhang},
  journal={2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition},
  year={2018},
  pages={1390-1399}
}
Motion representation plays a vital role in human action recognition in videos. In this study, we introduce a novel compact motion representation for video action recognition, named Optical Flow guided Feature (OFF), which enables the network to distill temporal information through a fast and robust approach. The OFF is derived from the definition of optical flow and is orthogonal to the optical flow. The derivation also provides theoretical support for using the difference between two frames… Expand

Figures, Tables, and Topics from this paper

PAN: Persistent Appearance Network with an Efficient Motion Cue for Fast Action Recognition
PAN: Towards Fast Action Recognition via Learning Persistence of Appearance
Spatiotemporal Relation Networks for Video Action Recognition
MotionSqueeze: Neural Motion Feature Learning for Video Understanding
MARS: Motion-Augmented RGB Stream for Action Recognition
Self-supervised Motion Representation via Scattering Local Motion Cues
Accurate Single-Stream Action Detection in Real-Time
...
1
2
3
4
5
...

References

SHOWING 1-10 OF 56 REFERENCES
Real-Time Action Recognition with Enhanced Motion Vector CNNs
Spatiotemporal Residual Networks for Video Action Recognition
Dynamic Image Networks for Action Recognition
Efficient Two-Stream Motion and Appearance 3D CNNs for Video Classification
Spatio-Temporal Vector of Locally Max Pooled Features for Action Recognition in Videos
Spatiotemporal Pyramid Network for Video Action Recognition
Deep Temporal Linear Encoding Networks
Action Recognition with Improved Trajectories
  • Heng Wang, C. Schmid
  • Mathematics, Computer Science
  • 2013 IEEE International Conference on Computer Vision
  • 2013
Temporal Segment Networks: Towards Good Practices for Deep Action Recognition
...
1
2
3
4
5
...