PAT++: a cautionary tale about generative visual augmentation for Object Re-identification

2025/07/19 by Leonardo Santiago Benitez Pereira, Pereira, Leonardo Santiago Benitez, Arathy Jeevan +1
Computer Science · Engineering · #Advanced Image and Video Retrieval Techniques #CCD and CMOS Imaging Sensors #Computer Vision and Pattern Recognition (cs.CV) #FOS: Computer and information sciences #Visual Attention and Saliency Detection

paper · pdf · doi:10.48550/arxiv.2507.15888

openalex publication_date 2025/07/19 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

Abstract

Generative data augmentation has demonstrated gains in several vision tasks, but its impact on object re-identification - where preserving fine-grained visual details is essential - remains largely unexplored. In this work, we assess the effectiveness of identity-preserving image generation for object re-identification. Our novel pipeline, named PAT++, incorporates Diffusion Self-Distillation into the well-established Part-Aware Transformer. Using the Urban Elements ReID Challenge dataset, we conduct extensive experiments with generated images used for both model training and query expansion. Our results show consistent performance degradation, driven by domain shifts and failure to retain identity-defining features. These findings challenge assumptions about the transferability of generative models to fine-grained recognition tasks and expose key limitations in current approaches to visual augmentation for identity-preserving applications.

Citations

• Diffusion Self-Distillation for Zero-Shot Customized Image Generation
• Score Forgetting Distillation: A Swift, Data-Free Method for Machine Unlearning in Diffusion Models
• Advances in Diffusion Models for Image Data Augmentation: A Review of Methods, Models, Evaluation Metrics and Future Research Directions
• Automated Virtual Product Placement and Assessment in Images using Diffusion Models
• InstantID: Zero-shot Identity-Preserving Generation in Seconds
• Part-Aware Transformer for Generalizable Person Re-identification
• DINOv2: Learning Robust Visual Features without Supervision
• GPT-4 Technical Report
• Adding Conditional Control to Text-to-Image Diffusion Models
• Adding Conditional Control to Text-to-Image Diffusion Models
• DreamBooth: Fine Tuning Text-to-Image Diffusion Models for Subject-Driven Generation
• DreamBooth: Fine Tuning Text-to-Image Diffusion Models for Subject-Driven Generation
• TransReID: Transformer-based Object Re-Identification
• Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks
• Re-ranking Person Re-identification with k-reciprocal Encoding
• End-to-end Learning of Deep Visual Representations for Image Retrieval

Related

• Other Tokens Matter: Exploring Global and Local Features of Vision Transformers for Object Re-Identification
• TransReID: Transformer-based Object Re-Identification
• Generalizable Object Re-Identification via Visual In-Context Prompting
• Generative Unlearning for Any Identity
• From Poses to Identity: Training-Free Person Re-Identification via Feature Centralization