{"ID":2849615,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2510.23301","arxiv_id":"2510.23301","title":"MDReID: Modality-Decoupled Learning for Any-to-Any Multi-Modal Object Re-Identification","abstract":"Real-world object re-identification (ReID) systems often face modality inconsistencies, where query and gallery images come from different sensors (e.g., RGB, NIR, TIR). However, most existing methods assume modality-matched conditions, which limits their robustness and scalability in practical applications. To address this challenge, we propose MDReID, a flexible any-to-any image-level ReID framework designed to operate under both modality-matched and modality-mismatched scenarios. MDReID builds on the insight that modality information can be decomposed into two components: modality-shared features that are predictable and transferable, and modality-specific features that capture unique, modality-dependent characteristics. To effectively leverage this, MDReID introduces two key components: the Modality Decoupling Learning (MDL) and Modality-aware Metric Learning (MML). Specifically, MDL explicitly decomposes modality features into modality-shared and modality-specific representations, enabling effective retrieval in both modality-aligned and mismatched scenarios. MML, a tailored metric learning strategy, further enforces orthogonality and complementarity between the two components to enhance discriminative power across modalities. Extensive experiments conducted on three challenging multi-modality ReID benchmarks (RGBNT201, RGBNT100, MSVR310) consistently demonstrate the superiority of MDReID. Notably, MDReID achieves significant mAP improvements of 9.8\\%, 3.0\\%, and 11.5\\% in general modality-matched scenarios, and average gains of 3.4\\%, 11.8\\%, and 10.9\\% in modality-mismatched scenarios, respectively. The code is available at: \\textcolor{magenta}{https://github.com/stone96123/MDReID}.","short_abstract":"Real-world object re-identification (ReID) systems often face modality inconsistencies, where query and gallery images come from different sensors (e.g., RGB, NIR, TIR). However, most existing methods assume modality-matched conditions, which limits their robustness and scalability in practical applications. To address...","url_abs":"https://arxiv.org/abs/2510.23301","url_pdf":"https://arxiv.org/pdf/2510.23301v2","authors":"[\"Yingying Feng\",\"Jie Li\",\"Jie Hu\",\"Yukang Zhang\",\"Lei Tan\",\"Jiayi Ji\"]","published":"2025-10-27T13:08:46Z","proceeding":"cs.CV","tasks":"[\"cs.CV\"]","methods":"[]","has_code":false,"code_links":[{"ID":607726,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_id":2849615,"paper_url":"https://arxiv.org/abs/2510.23301","paper_title":"MDReID: Modality-Decoupled Learning for Any-to-Any Multi-Modal Object Re-Identification","repo_url":"https://github.com/stone96123/MDReID","is_official":false,"mentioned_in_paper":false,"mentioned_in_github":true,"github_stars":0}]}