Track papers across labs, topics, and trends.
Search papers, labs, and topics across Lattice.
This paper introduces ReSAGE-PAR, a novel pipeline that enhances Pedestrian Attribute Recognition (PAR) by addressing data scarcity through controlled image synthesis using diffusion models. The approach effectively bridges the domain gap between high-quality pre-training data and low-resolution surveillance images while ensuring reliable attribute verification to mitigate generative hallucinations. Experimental results show that integrating ReSAGE-PAR into PAR training leads to performance improvements of up to 8.7% on standard backbones, elevating state-of-the-art frameworks in the field.
Generating high-fidelity pedestrian images can significantly boost attribute recognition performance, overcoming the limitations of low-quality surveillance data.
To address the limited diversity and data scarcity in Pedestrian Attribute Recognition (PAR), we explore image synthesis using diffusion models guided by attribute-based prompts. While this enables the controlled generation of pedestrian images, it faces two critical challenges: (i) the domain gap between high-quality pre-training data and low-resolution, non-standard surveillance crops, and (ii) the need for reliable attribute verification to prevent generative hallucinations. In this paper, we introduce a robust generate-score-autolabel pipeline called ReSAGE-PAR (REpresentational Similarity Assessment for Generative Expansion in PAR) that bridges this domain gap and enables scalable, high-fidelity dataset expansion. First, we adapt pre-trained diffusion models to native PAR resolutions using a tailored LoRA-based Image-to-Image approach. Second, we extract vision-language alignment scores between the generated images and their conditioning prompts, utilizing a comprehensive prompting strategy that includes label-consistent and inconsistent complements. Finally, we formulate a Bayesian classifier that converts these continuous scores into reliable binary pseudo-labels. Extensive evaluations demonstrate the effectiveness of ReSAGE-PAR in preserving spatial priors and verifying attributes. When integrated into PAR training, ReSAGE-PAR consistently yields significant improvements-achieving gains of up to 8.7% on standard backbones and pushing state-of-the-art frameworks to new performance levels. This proves its value as an architecture-agnostic solution for scalable PAR enhancement. The complete codebase for ReSAGE-PAR is publicly available at http://www-vpu.eps.uam.es/publications/ReSAGE-PAR.