Identifying the key frames: An attention-aware sampling method for action recognition

doi:10.1016/j.patcog.2022.108797

CORC > 自动化研究所 > 中国科学院自动化研究所 > 多模态人工智能系统全国重点实验室

	Identifying the key frames: An attention-aware sampling method for action recognition
	Dong, Wenkai 1,2,4; Zhang, Zhaoxiang 1,2,3,4; Song, Chunfeng 1,2,4; Tan, Tieniu 1,2,4
刊名	PATTERN RECOGNITION
	2022-10-01
卷号	130 页码:11
关键词	Action recognition Deep learning Reinforcement learning Pseudo labels
ISSN号	0031-3203
DOI	10.1016/j.patcog.2022.108797
通讯作者	Zhang, Zhaoxiang(zhaoxiang.zhang@ia.ac.cn)
英文摘要	Deep learning based methods have achieved remarkable progress in action recognition. Existing works mainly focus on designing novel deep architectures to learn video representations for action recognition. Most existing methods treat sampled frames equally and average all the frame-level predictions to generate video-level predictions at the testing stage. However, within a video, discriminative actions may occur sparsely in a few frames whereas most other frames are irrelevant to the ground truth which may even lead to wrong results. As a result, we think that the strategy of selecting relevant frames would be a further important key to enhance the existing deep learning based action recognition. In this paper, we propose an attention-aware sampling method for action recognition, which aims to discard the irrelevant and misleading frames and preserve the most discriminative frames. We formulate the process of mining key frames from videos as a Markov decision process and train the attention agent through deep reinforcement learning without extra labels. The agent takes features and predictions from the baseline model as inputs and generates importance scores for all frames. Moreover, our approach is extensible, which can be applied to different existing deep learning based action recognition models. We achieve very competitive action recognition performance on two widely used action recognition datasets. (c) 2022 Elsevier Ltd. All rights reserved.
资助项目	Major Project for New Generation of AI[2018AAA0100400] ; National Natural Science Foundation of China[61836014] ; National Natural Science Foundation of China[U21B2042] ; National Natural Science Foundation of China[62006231] ; National Natural Science Foundation of China[62072457] ; National Youth Talent Support Program
WOS关键词	NETWORK
WOS研究方向	Computer Science ; Engineering
语种	英语
出版者	ELSEVIER SCI LTD
WOS记录号	WOS:001027089400007
资助机构	Major Project for New Generation of AI ; National Natural Science Foundation of China ; National Youth Talent Support Program
内容类型	期刊论文
源URL	[http://ir.ia.ac.cn/handle/173211/53669]
专题	多模态人工智能系统全国重点实验室
通讯作者	Zhang, Zhaoxiang
作者单位	1.Chinese Acad Sci CASIA, Inst Automat, Beijing, Peoples R China 2.Natl Lab Pattern Recognit NLPR, Beijing, Peoples R China 3.Chinese Acad Sci, Ctr Artificial Intelligence & Robot, HKISI, Beijing, Peoples R China 4.Univ Chinese Acad Sci UCAS, Beijing, Peoples R China
推荐引用方式 GB/T 7714	Dong, Wenkai,Zhang, Zhaoxiang,Song, Chunfeng,et al. Identifying the key frames: An attention-aware sampling method for action recognition[J]. PATTERN RECOGNITION,2022,130:11.
APA	Dong, Wenkai,Zhang, Zhaoxiang,Song, Chunfeng,&Tan, Tieniu.(2022).Identifying the key frames: An attention-aware sampling method for action recognition.PATTERN RECOGNITION,130,11.
MLA	Dong, Wenkai,et al."Identifying the key frames: An attention-aware sampling method for action recognition".PATTERN RECOGNITION 130(2022):11.