Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2603.14012 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866917342070439936 |
|---|---|
| author | Li, Jiachen Gong, Xiaojin Zhang, Dongping |
| author_facet | Li, Jiachen Gong, Xiaojin Zhang, Dongping |
| contents | Domain Generalized person Re-identification (DG Re-ID) is a challenging task, where models are trained on source domains but tested on unseen target domains. Although previous pure vision-based models have achieved significant progress, the performance remains further improved. Recently, Vision-Language Models (VLMs) present outstanding generalization capabilities in various visual applications. However, directly adapting a VLM to Re-ID shows limited generalization improvement. This is because the VLM only produces with global features that are insensitive to ID nuances. To tacle this problem, we propose a CLIP-based multi-grained vision-language alignment framework in this work. Specifically, several multi-grained prompts are introduced in language modality to describe different body parts and align with their counterparts in vision modality. To obtain fine-grained visual information, an adaptively masked multi-head self-attention module is employed to precisely extract specific part features. To train the proposed module, an MLLM-based visual grounding expert is employed to automatically generate pseudo labels of body parts for supervision. Extensive experiments conducted on both single- and multi-source generalization protocols demonstrate the superior performance of our approach. The implementation code will be released at https://github.com/RikoLi/MUVA. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2603_14012 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Multi-Grained Vision-Language Alignment for Domain Generalized Person Re-Identification Li, Jiachen Gong, Xiaojin Zhang, Dongping Computer Vision and Pattern Recognition Domain Generalized person Re-identification (DG Re-ID) is a challenging task, where models are trained on source domains but tested on unseen target domains. Although previous pure vision-based models have achieved significant progress, the performance remains further improved. Recently, Vision-Language Models (VLMs) present outstanding generalization capabilities in various visual applications. However, directly adapting a VLM to Re-ID shows limited generalization improvement. This is because the VLM only produces with global features that are insensitive to ID nuances. To tacle this problem, we propose a CLIP-based multi-grained vision-language alignment framework in this work. Specifically, several multi-grained prompts are introduced in language modality to describe different body parts and align with their counterparts in vision modality. To obtain fine-grained visual information, an adaptively masked multi-head self-attention module is employed to precisely extract specific part features. To train the proposed module, an MLLM-based visual grounding expert is employed to automatically generate pseudo labels of body parts for supervision. Extensive experiments conducted on both single- and multi-source generalization protocols demonstrate the superior performance of our approach. The implementation code will be released at https://github.com/RikoLi/MUVA. |
| title | Multi-Grained Vision-Language Alignment for Domain Generalized Person Re-Identification |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2603.14012 |