Video Action Recognition Based on Key-frame

Authors

Mingxiao Li, State Key Laboratory of Virtual Reality Technology and Systems, Beihang University, Beijing 100191, China;
Qichuan Geng, State Key Laboratory of Virtual Reality Technology and Systems, Beihang University, Beijing 100191, China;
Mo Hong, State Key Laboratory of Virtual Reality Technology and Systems, Beihang University, Beijing 100191, China;
Wu Wei, State Key Laboratory of Virtual Reality Technology and Systems, Beihang University, Beijing 100191, China;
Zhou Zhong, State Key Laboratory of Virtual Reality Technology and Systems, Beihang University, Beijing 100191, China;

Abstract

Abstract: Video action recognition is an important part of intelligent video analysis. In recent years, deep learning methods, especially the two-stream convolutional neural network achieved the state-of-the-art performance. However, most methods simply use uniform sampling to get frames, which may cause the loss of information in sampling interval. We propose a segmentation method and a key-frame extraction method for video action recognition, and combine them with a multi-temporal-scale two-stream network. Our framework achieves a 94.2% accuracy at UCF101 split1, which is the same as the state-of-the-art method’s performance.

Recommended Citation

Li, Mingxiao; Geng, Qichuan; Hong, Mo; Wei, Wu; and Zhong, Zhou (2019) "Video Action Recognition Based on Key-frame," Journal of System Simulation: Vol. 30: Iss. 7, Article 44.
DOI: 10.16182/j.issn1004731x.joss.201807044
Available at: https://dc-china-simulation.researchcommons.org/journal/vol30/iss7/44

First Page

2787

DOI Link

https://doi.org/10.16182/j.issn1004731x.joss.201807044

Last Page

2793

CLC

TP391.4

Recommended Citation

Li Mingxiao, Geng Qichuan, Mo Hong, Wu Wei, Zhou Zhong. Video Action Recognition Based on Key-frame[J]. Journal of System Simulation, 2018, 30(7): 2787-2793.

DOI

10.16182/j.issn1004731x.joss.201807044