logo

SCIENCE CHINA Information Sciences, Volume 63 , Issue 11 : 219102(2020) https://doi.org/10.1007/s11432-019-2782-3

Mobile person re-identification with a lightweight trident CNN

More info
  • ReceivedNov 8, 2019
  • AcceptedJan 8, 2020
  • PublishedApr 15, 2020

Abstract

There is no abstract available for this article.


Acknowledgment

This work was supported in part by National Natural Science Foundation of China (Grant Nos. 61772380, 61977027), Major Project for Technological Innovation of Hubei Province (Grant No. 2019AAA044), Science and Technology Major Project of Hubei Province (Next-Generation AI Technologies) (Grant No. 2019AEA170), Foundation for Innovative Research Groups of Hubei Province (Grant No. 2017CFA007), Science and Technology Planning Project of Shenzhen (Grant No. JCYJ20170818112550194), and Open Foundation for Engineering Research Center of Hubei Province for Clothing Information (Grant No. 184084012).


Supplement

Appendix A.


References

[1] Bai S, Bai X, Tian Q. Scalable Person Re-identification on Supervised Smoothed Manifold. In: Proceedings of IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2017. 2530--2539. Google Scholar

[2] Hou R, Ma B, Chang H, et al. VRSTC: occlusion-free video person re-identification. In: Proceedings of IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2019. 7183--7192. Google Scholar

[3] Bai S, Tang P, Torr P H S, et al. Re-ranking via metric fusion for object retrieval and person re-identification. In: Proceedings of IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2019. 740--749. Google Scholar

[4] Krizhevsky A, Sutskever I, Hinton G E. Imagenet classification with deep convolutional neural networks. In: Proceedings of Advances in Neural Information Processing Systems 25 (NIPS 2012), 2012. 1097--1105. Google Scholar

[5] Yu R, Dou Z-Y, Bai S, et al. Hard-aware point-to-set deep metric for person re-identification. In: Proceedings of the European Conference on Computer Vision (ECCV), 2018. 188--204. Google Scholar

[6] Yu R, Chao Z Z, Bai S, et al. Divide and fuse: a re-ranking approach for person re-identification. In: Proceedings of British Machine Vision Association, 2017. 1--13. Google Scholar

[7] Zhao C, Chen K, Zang D. Uncertainty-optimized deep learning model for small-scale person re-identification. Sci China Inf Sci, 2019, 62: 220102 CrossRef Google Scholar

[8] Chen D, Tang Y, Zhang H, et al. Incremental factorization of big time series data with blind factor approximation. IEEE Trans Knowl Data Eng, 2019. Google Scholar

[9] Huang T, Xu Y, Bai S. Feature context learning for human parsing. Sci China Inf Sci, 2019, 62: 220101 CrossRef Google Scholar

  • Figure 1

    (Color online) The solution of T-CNN architecture for person re-identification. The input images include the triplet samples denoted by $\langle~I,~I^{+},~I^{-}\rangle$; the T-CNN comprises the s-Net, c-Net, and b-Net; person feature ($R$) is formed via cascade learning and $l_{2}$-normalization; metric learning applies to the final results of person re-identification.

Copyright 2020  CHINA SCIENCE PUBLISHING & MEDIA LTD.  中国科技出版传媒股份有限公司  版权所有

京ICP备14028887号-23       京公网安备11010102003388号