
Kernel distance metric learning using pairwise constraints for person re-identification
- Author
- Bac Nguyen Cong (UGent) and Bernard De Baets (UGent)
- Organization
- Abstract
- Person re-identification is a fundamental task in many computer vision and image understanding systems. Due to appearance variations from different camera views, person re-identification still poses an important challenge. In the literature, KISSME has already been introduced as an effective distance metric learning method using pairwise constraints to improve the re-identification performance. Computationally, it only requires two inverse covariance matrix estimations. However, the linear transformation induced by KISSME is not powerful enough for more complex problems. We show that KISSME can be kernelized, resulting in a nonlinear transformation, which is suitable for many real-world applications. Moreover, the proposed kernel method can be used for learning distance metrics from structured objects without having a vectorial representation. The effectiveness of our method is validated on five publicly available data sets. To further apply the proposed kernel method efficiently when data are collected sequentially, we introduce a fast incremental version that learns a dissimilarity function in the feature space without estimating the inverse covariance matrices. The experiments show that the latter variant can obtain competitive results in a computationally efficient manner.
- Keywords
- Distance metric learning, dissimilarity learning, person re-identification, kernel-based learning, PEDESTRIAN RECOGNITION, CLASSIFICATION, SCALE, ALGORITHMS, FRAMEWORK, FEATURES
Downloads
-
(...).pdf
- full text
- |
- UGent only
- |
- |
- 1.39 MB
Citation
Please use this url to cite or link to this publication: http://hdl.handle.net/1854/LU-8575775
- MLA
- Nguyen Cong, Bac, and Bernard De Baets. “Kernel Distance Metric Learning Using Pairwise Constraints for Person Re-Identification.” IEEE TRANSACTIONS ON IMAGE PROCESSING, vol. 28, no. 2, 2019, pp. 589–600, doi:10.1109/tip.2018.2870941.
- APA
- Nguyen Cong, B., & De Baets, B. (2019). Kernel distance metric learning using pairwise constraints for person re-identification. IEEE TRANSACTIONS ON IMAGE PROCESSING, 28(2), 589–600. https://doi.org/10.1109/tip.2018.2870941
- Chicago author-date
- Nguyen Cong, Bac, and Bernard De Baets. 2019. “Kernel Distance Metric Learning Using Pairwise Constraints for Person Re-Identification.” IEEE TRANSACTIONS ON IMAGE PROCESSING 28 (2): 589–600. https://doi.org/10.1109/tip.2018.2870941.
- Chicago author-date (all authors)
- Nguyen Cong, Bac, and Bernard De Baets. 2019. “Kernel Distance Metric Learning Using Pairwise Constraints for Person Re-Identification.” IEEE TRANSACTIONS ON IMAGE PROCESSING 28 (2): 589–600. doi:10.1109/tip.2018.2870941.
- Vancouver
- 1.Nguyen Cong B, De Baets B. Kernel distance metric learning using pairwise constraints for person re-identification. IEEE TRANSACTIONS ON IMAGE PROCESSING. 2019;28(2):589–600.
- IEEE
- [1]B. Nguyen Cong and B. De Baets, “Kernel distance metric learning using pairwise constraints for person re-identification,” IEEE TRANSACTIONS ON IMAGE PROCESSING, vol. 28, no. 2, pp. 589–600, 2019.
@article{8575775, abstract = {{Person re-identification is a fundamental task in many computer vision and image understanding systems. Due to appearance variations from different camera views, person re-identification still poses an important challenge. In the literature, KISSME has already been introduced as an effective distance metric learning method using pairwise constraints to improve the re-identification performance. Computationally, it only requires two inverse covariance matrix estimations. However, the linear transformation induced by KISSME is not powerful enough for more complex problems. We show that KISSME can be kernelized, resulting in a nonlinear transformation, which is suitable for many real-world applications. Moreover, the proposed kernel method can be used for learning distance metrics from structured objects without having a vectorial representation. The effectiveness of our method is validated on five publicly available data sets. To further apply the proposed kernel method efficiently when data are collected sequentially, we introduce a fast incremental version that learns a dissimilarity function in the feature space without estimating the inverse covariance matrices. The experiments show that the latter variant can obtain competitive results in a computationally efficient manner.}}, author = {{Nguyen Cong, Bac and De Baets, Bernard}}, issn = {{1057-7149}}, journal = {{IEEE TRANSACTIONS ON IMAGE PROCESSING}}, keywords = {{Distance metric learning,dissimilarity learning,person re-identification,kernel-based learning,PEDESTRIAN RECOGNITION,CLASSIFICATION,SCALE,ALGORITHMS,FRAMEWORK,FEATURES}}, language = {{eng}}, number = {{2}}, pages = {{589--600}}, title = {{Kernel distance metric learning using pairwise constraints for person re-identification}}, url = {{http://doi.org/10.1109/tip.2018.2870941}}, volume = {{28}}, year = {{2019}}, }
- Altmetric
- View in Altmetric
- Web of Science
- Times cited: