Model improves person reidentification despite clothes changes
MINER: Modeling Inter- and Intra-Person Variance for Reidentification
Computer Vision and Pattern Recognition
Summary
Person reidentification is the task of recognizing the same person across different cameras or times, even if their appearance changes, such as wearing different clothes. The authors address this challenge by creating a system that balances the differences between people while reducing variation caused by outfit changes. Their approach, called MINER, uses a method that contrasts challenging examples to suppress the influence of clothes. Tests on several benchmarks show that MINER improves accuracy and performs very well compared to other methods.
What this means in practice
- •For security system developers: Improve identity tracking in surveillance systems when people change clothes frequently.
- •For retail analytics teams: Enhance customer behavior analysis across time and store areas despite outfit changes.
Authors
Oron Nir, Eliyahu Strugo, Ariel Shamir
Abstract
Clothes-Changing Person Re-Identification (CC-ReID) remains challenging due to drastic appearance shifts across outfits. Most prior works devise biometric signal-extraction heuristics to enhance inter-person variance, e.g., using segmentation masks, while leaving intra-person variance uncontrolled. We present MINER, a simple model-agnostic framework that suppresses garment bias by contrasting hard examples and adaptively balancing dataset-dependent intra- and inter-person variance. Extensive experiments show that MINER improves retrieval across CC-ReID benchmarks, achieving state-of-the-art performance on four datasets and ranking among the leaders in all three tracks of the open TVRID RGB-Depth ReID challenge.