Pose-Appearance Relational Modeling for Video Action Recognition
文献类型：期刊论文
| 作者 | Cui, Mengmeng2; Wang, Wei2 ; Zhang, Kunbo1,2; Sun, Zhenan1,2; Wang, Liang1,2
|
| 刊名 | IEEE TRANSACTIONS ON IMAGE PROCESSING
 |
| 出版日期 | 2023 |
| 卷号 | 32页码:295-308 |
| ISSN号 | 1057-7149 |
| 关键词 | Action recognition 2D pose-appearance relational modeling temporal attention LSTM |
| DOI | 10.1109/TIP.2022.3228156 |
| 通讯作者 | Wang, Wei(wangwei@nlpr.ia.ac.cn) |
| 英文摘要 | Recent studies of video action recognition can be classified into two categories: the appearance-based methods and the pose-based methods. The appearance-based methods generally cannot model temporal dynamics of large motion well by virtue of optical flow estimation, while the pose-based methods ignore the visual context information such as typical scenes and objects, which are also important cues for action understanding. In this paper, we tackle these problems by proposing a Pose-Appearance Relational Network (PARNet), which models the correlation between human pose and image appearance, and combines the benefits of these two modalities to improve the robustness towards unconstrained real-world videos. There are three network streams in our model, namely pose stream, appearance stream and relation stream. For the pose stream, a Temporal Multi-Pose RNN module is constructed to obtain the dynamic representations through temporal modeling of 2D poses. For the appearance stream, a Spatial Appearance CNN module is employed to extract the global appearance representation of the video sequence. For the relation stream, a Pose-Aware RNN module is built to connect pose and appearance streams by modeling action-sensitive visual context information. Through jointly optimizing the three modules, PARNet achieves superior performances compared with the state-of-the-arts on both the pose-complete datasets (KTH, Penn-Action, UCF11) and the challenging pose-incomplete datasets (UCF101, HMDB51, JHMDB), demonstrating its robustness towards complex environments and noisy skeletons. Its effectiveness on NTU-RGBD dataset is also validated even compared with 3D skeleton-based methods. Furthermore, an appearance-enhanced PARNet equipped with a RGB-based I3D stream is proposed, which outperforms the Kinetics pre-trained competitors on UCF101 and HMDB51. The better experimental results verify the potentials of our framework by integrating various modules. |
| WOS关键词 | ATTENTION NETWORK ; LSTM |
| 资助项目 | National Natural Science Foundation of China[61976214] ; National Natural Science Foundation of China[62071468] ; National Natural Science Foundation of China[62006225] ; National Natural Science Foundation of China[61806197] ; National Natural Science Foundation of China[6207146] ; Strategic Priority Research Program of Chinese Academy of Sciences[XDA27040700] ; Beijing Municipal Natural Science Foundation[4214075] |
| WOS研究方向 | Computer Science ; Engineering |
| 语种 | 英语 |
| 出版者 | IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC |
| WOS记录号 | WOS:000902111900021 |
| 资助机构 | National Natural Science Foundation of China ; Strategic Priority Research Program of Chinese Academy of Sciences ; Beijing Municipal Natural Science Foundation |
| 源URL | [http://ir.ia.ac.cn/handle/173211/51084]  |
| 专题 | 多模态人工智能系统全国重点实验室 |
| 通讯作者 | Wang, Wei |
| 作者单位 | 1.Univ Chinese Acad Sci, Beijing 101408, Peoples R China 2.Chinese Acad Sci, Inst Automat, Ctr Res Intelligent Percept & Comp, Natl Lab Pattern Recognit, Beijing 100190, Peoples R China |
| 推荐引用方式 GB/T 7714 | Cui, Mengmeng,Wang, Wei,Zhang, Kunbo,et al. Pose-Appearance Relational Modeling for Video Action Recognition[J]. IEEE TRANSACTIONS ON IMAGE PROCESSING,2023,32:295-308. |
| APA | Cui, Mengmeng,Wang, Wei,Zhang, Kunbo,Sun, Zhenan,&Wang, Liang.(2023).Pose-Appearance Relational Modeling for Video Action Recognition.IEEE TRANSACTIONS ON IMAGE PROCESSING,32,295-308. |
| MLA | Cui, Mengmeng,et al."Pose-Appearance Relational Modeling for Video Action Recognition".IEEE TRANSACTIONS ON IMAGE PROCESSING 32(2023):295-308. |
入库方式： OAI收割
来源：自动化研究所
浏览0
下载0
收藏0
其他版本
除非特别说明，本系统中所有内容都受版权保护，并保留所有权利。