Pedestrian Attributes Recognition in Surveillance Scenarios Using Multi-Task Lightweight Convolutional Neural Network
Source abstract
Abstract
Pedestrian attributes (such as gender, age, hairstyle, and clothing) can effectively represent the appearance of pedestrians. These are high-level semantic features that are robust to illumination, deformation, etc. Therefore, they can be widely used in person re-identification, video structuring analysis and other applications. In this paper, a pedestrian attributes recognition method for surveillance scenarios using a multi-task lightweight convolutional neural network is proposed. Firstly, the labels of the attributes for each pedestrian image are integrated into a label vector. Then, a multi-task lightweight Convolutional Neural Network (CNN) is designed, which consists of five convolutional layers, three pooling layers and two fully connected layers to extract the deep features of pedestrian images. Considering that the data distribution of the datasets is unbalanced, the loss function is improved based on the sigmoid cross-entropy, and the scale factor is added to balance the amount of various attributes data. Through training the network, the mapping relationship model between the deep features of pedestrian images and the integration label vector of their attributes is established, which can be used to predict each attribute of the pedestrian. The experiments were conducted on two public pedestrian attributes datasets in surveillance scenarios, namely PETA and RAP. The results show that, compared with the state-of-the-art pedestrian attributes recognition methods, the proposed method can achieve a superior accuracy by 91.88% on PETA and 87.44% on RAP respectively.
Source: crossref · Quality: baseline
Research context
OpenAlex topics: Video Surveillance and Tracking Methods · Fire Detection and Safety Systems · Advanced Neural Network Applications
SciVX discovery tags are terminology-based. OpenAlex topics are source-provided classifications.
Affiliations
Faculty of Information Technology, Beijing University of Technology, Beijing 100124, China
Beijing Key Laboratory of Computational Intelligence and Intelligent System, Beijing University of Technology, Beijing 100124, China
Funding
National Natural Science Foundation of China · Beijing Municipal Natural Science Foundation Cooperation Beijing Education Committee
References (19)
Show source-provided references
1. Wang, X., Zheng, S., Yang, R., Luo, B., and Tang, J. (2019). Pedestrian Attribute Recognition: A Survey. arXiv.
2. Layne, R., Hospedales, T.M., and Gong, S. (2014). Attributes-Based Re-Identification, Springer. Person Re-Identification.10.1007/978-1-4471-6296-4_5 ↗
3. Zhu, J., Liao, S., Lei, Z., Yi, D., and Li, S.Z. (2013, January 2–8). Pedestrian Attribute Classification in Surveillance: Database and Evaluation. Proceedings of the IEEE International Conference on Computer Vision (ICCV), Sydney, NSW, Australia.10.1109/ICCVW.2013.51 ↗
4. Schumann, A., and Monari, E. (2014, January 26–29). A soft-biometrics dataset for person tracking and re-identification. Proceedings of the 2014 11th IEEE International Conference on Advanced Video and Signal Based Surveillance (AVSS), Seoul, Korea.10.1109/AVSS.2014.6918667 ↗
5. Sudowe, P., Spitzer, H., and Leibe, B. (2015, January 7–13). Person attribute recognition with a jointly-Trained holistic CNN model. Proceedings of the IEEE International Conference on Computer Vision (ICCV), Santiago, Chile.10.1109/ICCVW.2015.51 ↗
6. Pedestrian attributes recognition in surveillance scenarios with hierarchical multi-task CNN models
7. Wang, J., Zhu, X., Gong, S., and Li, W. (2017, January 22–29). Attribute Recognition by Joint Recurrent Learning of Context and Correlation. Proceedings of the 2017 IEEE International Conference on Computer Vision (ICCV), Venice, Italy.10.1109/ICCV.2017.65 ↗
8. Zhou, Y., Yu, K., Leng, B., Zhang, Z., Li, D., Huang, K., Feng, B., and Yao, C. (2017). Weakly-supervised learning of mid-level features for pedestrian attribute recognition and localization. arXiv.10.5244/C.31.69 ↗
9. Sarafianos, N., Giannakopoulos, T., Nikou, C., and Kakadiaris, I.A. (2017, January 22–29). Curriculum learning for multi-task classification of visual attributes. Proceedings of the 2017 IEEE International Conference on Computer Vision Workshop, Venice, Italy.10.1109/ICCVW.2017.306 ↗
10. Lin, Y., Zheng, L., Zheng, Z., Wu, Y., and Yang, Y. (2017). Improving person re-identification by attribute and identity learning. arXiv.
11. A discriminatively learned CNN embedding for person re-identification10.1145/3159171 ↗
12. Attention-Based Pedestrian Attribute Analysis10.1109/TIP.2019.2919199 ↗
13. Hassanzadeh, H.R., and Rouhani, M. (2010, January 28–30). A Multi-Objective Gravitational Search Algorithm. Proceedings of the 2012 2nd International Conference on Computational Intelligence, Communication Systems and Networks, CICSyN, Liverpool, UK.10.1109/CICSyN.2010.32 ↗
14. Kendall, A., Gal, Y., and Cipolla, R. (2018, January 18–23). Multi-Task Learning Using Uncertainty to Weigh Losses for Scene Geometry and Semantics. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Salt Lake City, UT, USA.
15. Deng, Y., Luo, P., Loy, C.C., and Tang, X. (2014, January 3–7). Pedestrian attribute recognition at far distance. Proceedings of the 22nd ACM International Conference on Multimedia, Orlando, FL, USA.10.1145/2647868.2654966 ↗
16. Li, D., Zhang, Z., Chen, X., Ling, H., and Huang, K. (2016). Arichly annotated dataset for pedestrian attribute recognition. Computer Vision and Pattern Recognition. arXiv.
17. Multi-label convolutional neural network based pedestrian attribute classification10.1016/j.imavis.2016.07.004 ↗
18. Gray, D., and Tao, H. (2008). Viewpoint invariant pedestrian recognition with an ensemble of localized Features. Lecture Notes in Computer Science, Proceedings of the European Conference on Computer Vision (ECCV), Marseille, France, 12–18 October 2008, Springer.10.1007/978-3-540-88682-2_21 ↗
19. Li, D., Chen, X., and Huang, K. (2015, January 3–6). Multi-attribute learning for pedestrian attribute recognition in surveillance scenarios. Proceedings of the 2015 3rd IAPR Asian Conference on Pattern Recognition (ACPR), Kuala Lumpur, Malaysia.10.1109/ACPR.2015.7486476 ↗