时空特征融合深度学习网络人体行为识别方法
Action recognition method of spatio-temporal feature fusion deep learning network
查看参考文献14篇
文摘
|
基于自然场景图像的人体行为识别方法中遮挡、背景干扰、光照不均匀等因素影响识别结果,利用人体三维骨架序列的行为识别方法可以克服上述缺点。首先,考虑人体行为的时空特性,提出一种时空特征融合深度学习网络人体骨架行为识别方法;其次,根据骨架几何特征建立视角不变性特征表示,CNN(Convolutional Neural Network)网络学习骨架的局部空域特征,作用于空域的LSTM(Long Short Term Memory)网络学习骨架空域节点之间的相关性特征,作用于时域的LSTM网络学习骨架序列时空关联性特征;最后,利用NTU RGB+D数据库验证文中算法。实验结果表明:算法识别精度有所提高,对于多视角骨架具有较强的鲁棒性。 |
其他语种文摘
|
Action recognition from natural scene was affected by complex illumination conditions and cluttered backgrounds. There was a growing interest in solving these problems by using 3D skeleton data. Firstly, considering the spatio-temporal features of human actions, a spatio-temporal fusion deep learning network for action recognition was proposed; Secondly, view angle invariant character was constructed based on geometric features of the skeletons. Local spatial character was extracted by short -time CNN networks. A spatio -LSTM network was used to learn the relation between joints of a skeleton frame. Temporal LSTM was used to learn spatio -temporal relation between skeleton sequences. Lastly, NTU RGB+D datasets were used to evaluate this network, the network proposed achieved the state-of-the-art performance for 3D human action analysis. Experimental results show that this network has strong robustness for view invariant sequences. |
来源
|
红外与激光工程
,2018,47(2):0203007-1-0203007-6 【核心库】
|
DOI
|
10.3788/IRLA201847.0203007
|
关键词
|
时空特征
;
融合
;
骨架
;
视角不变
|
地址
|
1.
辽宁石油化工大学信息与控制工程学院, 辽宁, 抚顺, 113001
2.
中国科学院沈阳自动化研究所, 机器人学国家重点实验室, 辽宁, 沈阳, 110016
|
语种
|
中文 |
文献类型
|
研究性论文 |
ISSN
|
1007-2276 |
学科
|
自动化技术、计算机技术 |
基金
|
国家自然科学基金
|
文献收藏号
|
CSCD:6207499
|
参考文献 共
14
共1页
|
1.
Wang Jiang. Mining actionlet ensemble for action recognition with depth cameras.
IEEE Conference on Computer Vision and Pattern Recognition,2012:1290-1297
|
被引
10
次
|
|
|
|
2.
Luvizon D C. Learning features combination for human action recognition from skeleton sequences.
Pattern Recognition Letters,2017,99(11):13-20
|
被引
9
次
|
|
|
|
3.
Ji Xiaopeng. The spatial laplacian and temporal energy pyramid representation for human action recognition using depth sequence.
Knowledge-Based System,2017,122:64-74
|
被引
2
次
|
|
|
|
4.
Zhang Pengfei. View adaptive recurrent neural networks for high performance human action recognition from skeleton data.
ICCV 2017. International Conference on Computer Vision,2017:2136-2145
|
被引
1
次
|
|
|
|
5.
Du Yong. Hierarchical recurrent neural network for skeleton based action recognition.
IEEE Conference on Computer Vision and Pattern Recognition,2015:1110-1118
|
被引
5
次
|
|
|
|
6.
Vivek Veeriah. Differential recurrent neural networks for action recognition.
IEEE Conference on Computer Vision and Pattern Recognition,2015:4041-4049
|
被引
1
次
|
|
|
|
7.
Zhu Wentao. Co-occurrence feature learning for skeleton based action recognition using regularized deep LSTM networks.
AAAI,2016:3697-3704
|
被引
1
次
|
|
|
|
8.
Amir Shahroudy. NTU RGB +D: A large scale dataset for 3D human activity analysis.
IEEE Conference on Computer Vision and Pattern Recognition,2016:1010-1019
|
被引
8
次
|
|
|
|
9.
Liu Jun. Spatio-temporal LSTM with trust gates for 3D human action recognition.
IEEE Conference on Computer Vision and Pattern Recognition,2016
|
被引
1
次
|
|
|
|
10.
Liu Jun. Skeleton based human action recognition with glo bal context-aware attention LSTM networks.
IEEE Conference on Computer Vision and Pattern Recognition,2017:3671-3680
|
被引
2
次
|
|
|
|
11.
Huang Zhiwu. Deep learning on Lie groups for skeleton-based action recognitio.
IEEE Conference on Computer Vision and Pattern Recognition,2017:1243-1252
|
被引
2
次
|
|
|
|
12.
周飞燕. 卷积神经网络研究综述.
计算机学报,2017,40(6):1229-1250
|
被引
538
次
|
|
|
|
13.
罗海波. 基于深度学习的目标跟踪方法研究现状与展望.
红外与激光工程,2017,46(5):0502002
|
被引
18
次
|
|
|
|
14.
绍春艳. 采用高维数据聚类的目标跟踪.
红外与激光工程,2016,45(4):0428002
|
被引
1
次
|
|
|
|
|