Competitive token routing improves object re-identification accuracy
FM-ReID: Selective Competitive Token Routing for Object Re-Identification
Computer Vision and Pattern RecognitionArtificial Intelligence
Summary
Object re-identification means recognizing the same object, like a person or vehicle, in different images. This can be tricky when different objects look very similar overall, but have small unique details. The authors introduce FM-ReID, a new method that uses multiple focused queries to pick out these small details from the image’s data. This way, the method keeps important local clues separate instead of mixing everything into one big description. Their approach works well for recognizing animals, people, and vehicles across various tests.
What this means in practice
- •For surveillance system developers: Improve person re-identification in video feeds by detecting subtle clothing and accessory differences using competitive token selection.
- •For wildlife monitoring teams: Enhance identification of individual animals in camera trap images through detailed recognition of unique markings and scars.
Authors
Zhiqi Li, Xiaowei Zhou, Zeyuan Sun, Feng Gao, Junyu Dong
Abstract
Object re-identification (ReID) faces a recurring challenge: different identities can share highly similar global appearances, while the cues that distinguish them are localized, heterogeneous, and visible only under particular viewpoints. This challenge arises in animal ReID through markings, contours, and scars, in person ReID through subtle clothing and accessory cues, and in vehicle ReID through localized appearance details. Although visual foundation models encode such information in dense tokens, a single holistic descriptor can obscure discriminative local signals. We propose FM-ReID, an end-to-end framework that formulates local representation learning as selective competitive token routing. Its Competitive Fine-grained Mining module uses multiple mining queries and a residual query to compete for dense DINOv3 tokens. Above-prior selection retains tokens preferentially allocated to each mining query, while the residual slot receives tokens excluded from the retrieval descriptors. The resulting multi-query descriptors are jointly trained with a holistic representation for retrieval, without fixed spatial partitions or equal-area constraints. FM-ReID achieves strong results on animal, person, and vehicle ReID benchmarks, supporting competitive token routing as an effective way to augment holistic foundation-model representations.