Generating Pedestrian Images for Person Re-identification
Person re-identification (re-ID) is mainly used to search the target pedestrian in different cameras. In this paper, we employ generative adversarial network (GAN) to expand training samples and evaluate the performance of two different label assignment strategies for the generated samples. We also investigate how the number of generated samples influences the re-ID performance. We do several experiments on the Market1501 database, and the experimental results are of essential reference value to this research field.
KeywordsPerson re-identification GAN Generated samples
This work was supported by National Natural Science Foundation of China under Grant No. 61501327 and No. 61711530240, Natural Science Foundation of Tianjin under Grant No. 17JCZDJC30600 and No. 15JCQNJC01700, the Fund of Tianjin Normal University under Grant No.135202RC1703, the Open Projects Program of National Laboratory of Pattern Recognition under Grant No. 201700001 and No. 201800002, the China Scholarship Council No. 201708120039 and No. 201708120040, and the Tianjin Higher Education Creative Team Funds Program.
- 1.Zhang Z, Wang C, Xiao B, Zhou W, Liu S, Shi C. Cross-view action recognition via a continuous virtual path. In: IEEE conference on computer vision and pattern recognition. Portland; 2013. p. 2690–7.Google Scholar
- 3.Liao S, Hu Y, Zhu X, Li ZS. Person re-identification by local maximal occurrence representation and metric learning. In: IEEE conference on computer vision and pattern recognition. Boston; 2015. p. 2197–206.Google Scholar
- 4.Koestinger M, Hirzer M, Wohlhart P, Peter M, Horst B. Large scale metric learning from equivalence constraints. In: IEEE conference on computer vision and pattern recognition. Providence; 2012. p. 2288–95.Google Scholar
- 8.Zhang Z, Si T. Learning deep features from body and parts for person re-identification in camera networks. EURASIP J Wirel Commun Network. 2018;52.Google Scholar
- 10.Sun Y, Zheng L, Yang Y, Tian Q, Wang S. Beyond part models: person retrieval with refined part pooling; 2017. arXiv preprint arXiv:1711.09349.
- 11.Goodfellow I, Pouget-Abadie J, Mirza M, Xu B, Warde-Farley D, Ozair S, Courville A, Bengio Y. Generative adversarial nets. In: Advances in neural information processing systems. Montreal; 2014. p. 2672–80.Google Scholar
- 12.Zheng Z, Zheng L, Yang Y. Unlabeled samples generated by Gan improve the person re-identification baseline in vitro; 2017. arXiv preprint arXiv:1701.07717.
- 13.Zhong Z, Zheng L, Zheng Z, Li S, Yang Y. Camera style adaptation for person re-identification. In: IEEE conference on computer vision and pattern recognition; 2018.Google Scholar
- 14.Deng W, Zheng L, Kang G, Yang Y, Ye Q, Jiao J. Image-image domain adaptation with preserved self-similarity and domain-dissimilarity for person re-identification; 2017. arXiv:1711.07027.
- 15.Radford A, Metz L, Chintala S. Unsupervised representation learning with deep convolutional generative adversarial networks; 2015. arXiv preprint arXiv:1511.06434.
- 16.Zheng L, Shen L, Tian L, Wang S, Wang J, Tian Q. Scalable person re-identification: a benchmark. In: IEEE international conference on computer vision. Chile; 2015. p. 1116–24.Google Scholar
- 17.He K, Zhang X, Ren S, Sun J. Deep residual learning for image recognition. In: IEEE conference on computer vision and pattern recognition. Las Vegas; 2016. p. 770–8.Google Scholar