2023/03/31 by Xiaofang Zhou, Yu-Jie Zhong, Zhou, Xiao +6 · 6 citations
Computer Science · Engineering · #Artificial intelligence #Biology #Class (philosophy) #Code (set theory) #Computer Vision and Pattern Recognition (cs.CV) #Computer science #Domain Adaptation and Few-Shot Learning #Exploit #FOS: Computer and information sciences #Identification (biology) #Machine learning #Object (grammar) #Pairwise comparison #Remote-Sensing Image Classification #Set (abstract data type) #Source code #Video Surveillance and Tracking Methods
paper · pdf · doi:10.48550/arxiv.2303.18247
published in arXiv (Cornell University) (Cornell University)
openalex publication_date 2023/03/31 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28
Object re-identification (ReID) aims to find instances with the same identity as the given probe from a large gallery. Pairwise losses play an important role in training a strong ReID network. Existing pairwise losses densely exploit each instance as an anchor and sample its triplets in a mini-batch. This dense sampling mechanism inevitably introduces positive pairs that share few visual similarities, which can be harmful to the training. To address this problem, we propose a novel loss paradigm termed Sparse Pairwise (SP) loss that only leverages few appropriate pairs for each class in a mini-batch, and empirically demonstrate that it is sufficient for the ReID tasks. Based on the proposed loss framework, we propose an adaptive positive mining strategy that can dynamically adapt to diverse intra-class variations. Extensive experiments show that SP loss and its adaptive variant AdaSP loss outperform other pairwise losses, and achieve state-of-the-art performance across several ReID benchmarks. Code is available at https://github.com/Astaxanthin/AdaSP.