ViTA-PAR: Visual and Textual Attribute Alignment with Attribute Prompting for Pedestrian Attribute Recognition

2025-06-02

AI Summary

In pedestrian attribute recognition (PAR), fine-grained local attributes—such as accessories—are challenging to localize accurately due to their highly variable spatial positions, limiting the performance of existing methods relying on fixed-region partitioning. To address this, we propose a cross-modal prompt alignment framework that jointly learns visual attribute prompts and person-attribute contextual text templates, enabling adaptive local feature extraction under global semantic guidance. By integrating multimodal prompt learning with vision-language feature alignment, our approach eliminates reliance on predefined regions and supports cross-regional, multi-scale modeling. Evaluated on four mainstream PAR benchmarks, our method achieves state-of-the-art performance while maintaining efficient inference. The code and pretrained models are publicly available.

Technology Category

• Computer Vision: Multi-modal Vision
• Natural Language Processing: Prompt Engineering / Prompting
• Machine Learning: Multimodal Learning

Application Category

• Search and Retrieval-Augmented AI: Personalized, context-aware and across-device search
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Graph Algorithms and Modeling for the Web: Algorithms and analysis for heterogeneous, signed, attributed, multi-relational, temporal, higher-order, and annotated Web-related graphs

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 performance, 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 robustness and accuracy, but these methods are often restricted to attribute classes within fixed horizontal regions, leading to degraded performance when attributes appear in varying or unexpected 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 visual 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 attribute 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.

Problem

• Enhancing pedestrian attribute recognition accuracy
• Addressing attribute localization variability issues
• Aligning visual and textual attribute features

Innovation

• Visual attribute prompts capture global-to-local semantics
• Learnable prompt template enriches textual embeddings
• Aligns visual and textual features for effective fusion