Proceedings of the Korea Information Processing Society Conference (한국정보처리학회:학술대회논문집)
- 2021.11a
- /
- Pages.878-879
- /
- 2021
- /
- 2005-0011(pISSN)
- /
- 2671-7298(eISSN)
DOI QR Code
FTSnet: A Simple Convolutional Neural Networks for Action Recognition
FTSnet: 동작 인식을 위한 간단한 합성곱 신경망
- Zhao, Yulan (Dept. of Computer Science and Engineering, Jeonbuk National University) ;
- Lee, Hyo Jong (Dept. of Computer Science and Engineering, Jeonbuk National University)
- Published : 2021.11.04
Abstract
Most state-of-the-art CNNs for action recognition are based on a two-stream architecture: RGB frames stream represents the appearance and the optical flow stream interprets the motion of action. However, the cost of optical flow computation is very high and then it increases action recognition latency. We introduce a design strategy for action recognition inspired by a two-stream network and teacher-student architecture. There are two sub-networks in our neural networks, the optical flow sub-network as a teacher and the RGB frames sub-network as a student. In the training stage, we distill the feature from the teacher as a baseline to train student sub-network. In the test stage, we only use the student so that the latency reduces without computing optical flow. Our experiments show that its advantages over two-stream architecture in both speed and performance.
Keywords