Journal
IEEE TRANSACTIONS ON MULTIMEDIA
Volume 24, Issue -, Pages 1665-1677Publisher
IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
DOI: 10.1109/TMM.2021.3069562
Keywords
Benchmark testing; Training; Feature extraction; Task analysis; Gallium nitride; Semantics; Pose estimation; Fine-grained difference; pairwise loss; person re-identification; representation learning; triplet loss
Categories
Funding
- National Natural Science Foundation of China [62067004]
- Beijing Natural Science Foundation [4202039]
- Jiangxi Research Institute of Beihang University
Ask authors/readers for more resources
Person Re-Identification aims to re-identify individuals from different viewpoints using fine-grained appearance differences. A novel pairwise loss function is introduced to enable learning of fine-grained features by penalizing small differences exponentially and large differences moderately. Experimental results show that the proposed loss outperforms popular loss functions and enhances data efficiency.
Person Re-IDentification (ReID) aims at re-identifying persons from different viewpoints across multiple cameras. Capturing the fine-grained appearance differences is often the key to accurate person ReID, because many identities can be differentiated only when looking into these fine-grained differences. However, most state-of-the-art person ReID approaches, typically driven by a triplet loss, fail to effectively learn the fine-grained features as they are focused more on differentiating large appearance differences. To address this issue, we introduce a novel pairwise loss function that enables ReID models to learn the fine-grained features by adaptively enforcing an exponential penalization on the images of small differences and a bounded penalization on the images of large differences. The proposed loss is generic and can be used as a plugin to replace the triplet loss to significantly enhance different types of state-of-the-art approaches. Experimental results on four benchmark datasets show that the proposed loss substantially outperforms a number of popular loss functions by large margins; and it also enables significantly improved data efficiency.
Authors
I am an author on this paper
Click your name to claim this paper and add it to your profile.
Reviews
Recommended
No Data Available