Large margin relative distance learning for person re-identification
- Author(s): Husheng Dong 1, 2 ; Shengrong Gong 1, 3 ; Chunping Liu 1 ; Yi Ji 1 ; Shan Zhong 3
-
-
View affiliations
-
Affiliations:
1:
School of Computer Science and Technology , Soochow University , Suzhou , People's Republic of China ;
2: Suzhou Institute of Trade and Commerce , Suzhou , People's Republic of China ;
3: Changshu Institute of Science and Technology , Changshu , People's Republic of China
-
Affiliations:
1:
School of Computer Science and Technology , Soochow University , Suzhou , People's Republic of China ;
- Source:
Volume 11, Issue 6,
September
2017,
p.
455 – 462
DOI: 10.1049/iet-cvi.2016.0265 , Print ISSN 1751-9632, Online ISSN 1751-9640
Distance metric learning has achieved great success in person re-identification. Most existing methods that learn metrics from pairwise constraints suffer the problem of imbalanced data. In this study, the authors present a large margin relative distance learning (LMRDL) method which learns the metric from triplet constraints, so that the problem of imbalanced sample pairs can be bypassed. Different from existing triplet-based methods, LMRDL employs an improved triplet loss that enforces penalisation on the triplets with minimal inter-class distance, and this leads to a more stringent constraint to guide the learning. To suppress the large variations of pedestrian's appearance in different camera views, the authors propose to learn the metric over the intra-class subspace. The proposed method is formulated as a logistic metric learning problem with positive semi-definite constraint, and the authors derive an efficient optimisation scheme to solve it based on the accelerated proximal gradient approach. Experimental results show that the proposed method achieves state-of-the-art performance on three challenging datasets (VIPeR, PRID450S, and GRID).
Inspec keywords: pedestrians; image matching; learning (artificial intelligence); gradient methods
Other keywords: imbalanced data; proximal gradient approach; imbalanced sample pairs; large margin relative distance learning; person reidentification; triplet constraints; LMRDL; logistic metric learning problem; optimisation scheme; distance metric learning
Subjects: Image recognition; Optimisation techniques; Optimisation techniques; Knowledge engineering techniques; Computer vision and image processing techniques
References
-
-
1)
-
20. Yang, Y., Yang, J., Yan, J., et al: ‘Salient color names for person re-identification’. Proc. of European Conf. on Computer Vision, 2014, pp. 536–551.
-
-
2)
-
21. Liao, S., Hu, Y., Zhu, X., et al: ‘Person re-identification by local maximal occurrence representation and metric learning’. Proc. of Computer Vision and Pattern Recognition, 2015, pp. 2197–2206.
-
-
3)
-
27. Luo, P., Wang, X., Tang, X.: ‘Pedestrian parsing via deep decompositional network’. Proc. of Computer Vision and Pattern Recognition, 2013, pp. 2648–2655.
-
-
4)
-
1. Hirzer, M., Roth, P.M., Bischof, H.: ‘Person re-identification by efficient impostor-based metric learning’. Proc. of IEEE Conf. on Advanced Video and Signal-Based Surveillance, 2012, pp. 203–208.
-
-
5)
-
16. Farenzena, M., Bazzani, L., Perina, A., et al: ‘Person re-identification by symmetry-driven accumulation of local features’. Proc. of Computer Vision and Pattern Recognition, 2010, pp. 2360–2367.
-
-
6)
-
30. Zhang, Y., Li, B., Lu, H., et al: ‘Sample-specific SVM learning for person re-identification’. Proc. of Computer Vision and Pattern Recognition, 2016, pp. 1278–1287.
-
-
7)
-
13. Doretto, G., Sebastian, T., Tu, P., et al: ‘Appearance-based person reidentification in camera networks: problem overview and current approaches’, J. Ambient Intell. Human Comput., 2011, 2, (2), pp. 127–151.
-
-
8)
-
2. Köstinger, M., Hirzer, M., Wohlhart, R.P.M., et al: ‘Large scale metric learning from equivalence constraints’. Proc. of Computer Vision and Pattern Recognition, 2012, pp. 2288–2295.
-
-
9)
-
5. Mignon, A., Jurie, F.: ‘PCCA: A new approach for distance learning from sparse pairwise constraints’. Proc. of Computer Vision and Pattern Recognition, 2012, pp. 2666–2672.
-
-
10)
-
28. Zhang, L., Xiang, T., Gong, S.: ‘Learning a discriminative null space for person re-identification’. Proc. of Computer Vision and Pattern Recognition, 2016, pp. 1239–1248.
-
-
11)
-
18. Kviatkovsky, I., Adam, A., Rivlin, E.: ‘Color invariants for person reidentification’, IEEE Trans. Pattern Anal. Mach. Intell., 2013, 35, (7), pp. 1622–1634.
-
-
12)
-
36. Liu, X., Wang, H., Wu, Y., et al: ‘An ensemble color model for human re-identification’. Proc. of Applications of Computer Vision, 2015, pp. 868–875.
-
-
13)
-
9. Weinberger, K.Q., Saul, L.K.: ‘Distance metric learning for large margin nearest neighbor classification’, J. Mach. Learn. Res., 2009, 10, pp. 207–244.
-
-
14)
-
26. Chen, C.L., Xiang, T., Gong, S.: ‘Multi-camera activity correlation analysis’. Proc. of Computer Vision and Pattern Recognition, 2009, pp. 1988–1995.
-
-
15)
-
14. Vezzani, R., Baltieri, D., Cucchiara, R.: ‘People reidentification in surveillance and forensics:a survey’, ACM Comput. Surv., 2013, 46, (2), pp. 1–37.
-
-
16)
-
15. Henggeler, S.W., Schoenwald, S.K., Liao, J.G., et al: ‘A survey of approaches and trends in person re-identification’, Image Vis. Comput., 2014, 32, (4), pp. 270–286.
-
-
17)
-
24. Gray, D., Tao, H.: ‘Viewpoint invariant pedestrian recognition with an ensemble of localized features’. Proc. of European Conf. on Computer Vision, 2008, pp. 262–275.
-
-
18)
-
22. Xing, E.P., Ng, A.Y., Jordan, M.I., et al: ‘Distance metric learning, with application to clustering with side-information’, Adv. Neural Inf. Process. Syst., 2002, 15, pp. 505–512.
-
-
19)
-
23. Guillaumin, M., Verbeek, J., Schmid, C.: ‘Is that you? metric learning approaches for face identification’. Proc. of Int. Conf. on Computer Vision, 2009, pp. 498–505.
-
-
20)
-
3. Li, Z., Chang, S., Liang, F., et al: ‘Learning locally-adaptive decision functions for person verification’. Proc. of Computer Vision and Pattern Recognition, 2013, pp. 3610–3617.
-
-
21)
-
31. Xiao, T., Li, H., Ouyang, W., et al: ‘Learning deep feature representations with domain guided dropout for person re-identification’. Proc. of Computer Vision and Pattern Recognition, 2016.
-
-
22)
-
39. Chen, C.L., Liu, C., Gong, S.: ‘Person re-identification by manifold ranking’. Proc. of Int. Conf. on Image Processing, 2013, pp. 3567–3571.
-
-
23)
-
11. Parikh, N., Boyd, S.: ‘Proximal algorithms’, Found. Trends Optim., 2014, 1, (3), pp. 127–239.
-
-
24)
-
19. Ma, B., Su, Y., Jurie, F.: ‘Covariance descriptor based on bio-inspired features for person re-identification and face verification’, Image Vis. Comput., 2014, 32, (6), pp. 379–390.
-
-
25)
-
37. Chen, D., Yuan, Z., Hua, G., et al: ‘Similarity learning on an explicit polynomial kernel feature map for person re-identification’. Proc. of Computer Vision and Pattern Recognition, 2015, pp. 1565–1573.
-
-
26)
-
12. Tseng, P.: ‘On accelerated proximal gradient methods for convex-concave optimization’, submitted toSIAM J. Optim., 2008.
-
-
27)
-
34. Paisitkriangkrai, S., Wu, L., Shen, C., et al: ‘Structured learning of metric ensembles with application to person re-identification’, Comput. Vis. Image Underst., 2017, 156, pp. 51–65.
-
-
28)
-
33. Zhao, R., Ouyang, W., Wang, X.: ‘Learning mid-level filters for person re-identification’. Proc. of Computer Vision and Pattern Recognition, 2014, pp. 144–151.
-
-
29)
-
10. Cheng, D., Gong, Y., Zhou, S., et al: ‘Person re-identification by multi-channel parts-based cnn with improved triplet loss function’. Proc. of Computer Vision and Pattern Recognition, 2016, pp. 1335–1344.
-
-
30)
-
32. Zhao, R., Ouyang, W., Wang, X.: ‘Person re-identification by salience matching’. Proc. of Int. Conf. on Computer Vision, 2013, pp. 2528–2535.
-
-
31)
-
35. Chen, Y.C., Zheng, W.S., Lai, J.: ‘Mirror representation for modeling view-specific transform in person re-identification’. Proc. of Int. Conf. on Artificial Intelligence, 2015, pp. 3402–3408.
-
-
32)
-
29. Davis, J.V., Kulis, B., Jain, P., et al: ‘Information-theoretic metric learning’. Proc. of Int. Conf. on Machine learning, 2007, pp. 209–216.
-
-
33)
-
4. Liao, S., Li, S.Z.: ‘Efficient psd constrained asymmetric metric learning for person re-identification’. Proc. of Int. Conf. on Computer Vision, 2015, pp. 3685–3693.
-
-
34)
-
25. Roth, P.M., Hirzer, M., Köstinger, M., et al: ‘Mahalanobis distance learning for person re-identification’, in Gong, S. (Ed): ‘Person Re-Identification’ (Springer Press, 2014), pp. 247–267.
-
-
35)
-
17. Cheng, D.S., Cristani, M., Stoppa, M., et al: ‘Custom pictorial structures for re-identification’. Proc. of BMVC, 2011, pp. 68.1–68.11.
-
-
36)
-
8. Wang, F., Zuo, W., Lin, L., et al: ‘Joint learning of single-image and cross-image representations for person re-identification’. Proc. of Computer Vision and Pattern Recognition, 2016, pp. 1288–1296.
-
-
37)
-
7. Ding, S., Lin, L., Wang, G., et al: ‘Deep feature learning with relative distance comparison for person re-identification’, Pattern Recognit., 2015, 48, (10), pp. 2993–3003.
-
-
38)
-
6. Zheng, W.S., Gong, S., Xiang, T.: ‘Reidentification by relative distance comparison’, IEEE Trans. Pattern Anal. Mach. Intell., 2013, 35, (3), pp. 653–668.
-
-
39)
-
38. Ma, L., Yang, X., Tao, D.: ‘Person re-identification over camera networks using multi-task distance metric learning’, IEEE Trans. Image Process., 2014, 23, (8), pp. 3656–3670.
-
-
1)