A Deep and Structured Metric Learning Method for Robust Person Re-Identification. (December 2019)

• Record Type: Journal Article
• Authors: Ren, Chuan-Xian; Xu, Xiao-Lin; Lei, Zhen
• Is Part Of: Pattern recognition. Volume 96(2019:Dec.)
• Publication Date: 2019-12
• Subjects: Metric learning -- Feature extraction -- Deep neural networks -- Imbalance regularization -- Person re-identification; Pattern perception -- Periodicals; Perception des structures -- Périodiques; Patroonherkenning; 006.4
• DOI: 10.1016/j.patcog.2019.106995
• Languages: English
• ISSNs: 0031-3203

Abstract

Highlights: We propose a new method for deep metric learning and discriminant feature extraction. In formulating the loss function, the positive pairs of small distances and negative pairs of large distances are simultaneously removed to improve learning efficiency and prediction accuracy. To address the class-imbalance problem, a penalty factor is assigned to the negative pair distance. By using a linear function with the margin-based objectives, the weights update process is robust to large iterative loss values. The loss function is compatible with many network backbones. It thus can be used to induce new deep methods for metric learning. The method with the ResNet-50 backbone is used for person re-ID, and extensive experiments on benchmark datasets validate the effectiveness. Abstract: Person re-identification (re-ID) is to match different images of the same pedestrian. It has attracted increasing research interest in pattern recognition and machine learning. Traditionally, person re-ID is formulated as a metric learning problem with binary classification output. However, higher order relationship, such as triplet closeness among the instances, is ignored by such pair-wise based metric learning methods. Thus, the discriminative information hidden in these data is insufficiently explored. This paper proposes a new structured loss function to push the frontier of the person re-ID performance in realistic scenarios. The new loss function introduces two margin parameters. They operate as bounds to remove positive pairs of very small distances and negative pairs of large distances. A trade-off coefficient is assigned to the loss term of negative pairs to alleviate class-imbalance problem. By using a linear function with the margin-based objectives, the gradients w.r.t. weight matrices are no longer dependent on the iterative loss values in a multiplicative manner. This makes the weights update process robust to large iterative loss values. The new loss function is compatible with many deep learning architectures, thus, it induces new deep network with pair-pruning regularization for metric learning. To evaluate the performance of the proposed model, extensive experiments are conducted on benchmark datasets. The results indicate that the new loss together with the ResNet-50 backbone has excellent feature representation ability for person re-ID.