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

TL;DR

This paper critically evaluates the use of generative visual augmentation for object re-identification, revealing that current methods often degrade performance due to domain shifts and loss of identity-specific details.

Contribution

It introduces PAT++, a novel pipeline combining Diffusion Self-Distillation with Part-Aware Transformer to assess generative augmentation effects on re-identification.

Findings

Generative augmentation often causes performance degradation in object re-ID.

Domain shifts and loss of identity features are key issues in current methods.

Current generative models have limited transferability to fine-grained recognition tasks.

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.