[Project Page] [arXiv]

Abstract

The Pedestrian Attribute Recognition (PAR) task aims to identify various detailed attributes of an individual, such as clothing, accessories, and gender. To enhance PAR perfor- mance, a model must capture features ranging from coarse- grained global attributes (e.g., for identifying gender) to fine-grained local details (e.g., for recognizing accessories) that may appear in diverse regions. Recent research suggests that body part representation can enhance the model’s robust- ness and accuracy, but these methods are often restricted to attribute classes within fixed horizontal regions, leading to de- graded performance when attributes appear in varying or un- expected body locations. In this paper, we propose Visual and Textual Attribute Alignment with Attribute Prompting for Pedestrian Attribute Recognition, dubbed as ViTA-PAR, to enhance attribute recognition through specialized multimodal prompting and vision-language alignment. We introduce vi- sual attribute prompts that capture global-to-local semantics, enabling diverse attribute representations. To enrich textual embeddings, we design a learnable prompt template, termed person and attribute context prompting, to learn person and attributes context. Finally, we align visual and textual at- tribute features for effective fusion. ViTA-PAR is validated on four PAR benchmarks, achieving competitive performance with efficient inference. We release our code and model at https://github.com/mlnjeongpark/ViTA-PAR.

Code

The code will be released soon.

Citation

@article{park2025vita,
  title={ViTA-PAR: Visual and Textual Attribute Alignment with Attribute Prompting for Pedestrian Attribute Recognition},
  author={Park, Minjeong and Park, Hongbeen and Kim, Jinkyu},
  journal={arXiv preprint arXiv:2506.01411},
  year={2025}
}

---

For questions or inquiries, feel free to contact minjngpark@gmail.com