Diffusion-Based Data Augmentation for Image Recognition: A Systematic Analysis and Evaluation

TL;DR

This paper introduces UniDiffDA, a unified framework for diffusion-based data augmentation in image recognition, providing systematic analysis, fair benchmarking, and practical insights to improve low-data classification performance.

Contribution

It offers a comprehensive analytical framework and evaluation protocol for DiffDA, clarifying the design space and enabling fair comparison of methods.

Findings

DiffDA methods vary significantly in design and effectiveness.

The unified framework clarifies key differences and design choices.

Benchmarking reveals strengths and limitations of existing DiffDA strategies.

Abstract

Diffusion-based data augmentation (DiffDA) has emerged as a promising approach to improving classification performance under data scarcity. However, existing works vary significantly in task configurations, model choices, and experimental pipelines, making it difficult to fairly compare methods or assess their effectiveness across different scenarios. Moreover, there remains a lack of systematic understanding of the full DiffDA workflow. In this work, we introduce UniDiffDA, a unified analytical framework that decomposes DiffDA methods into three core components: model fine-tuning, sample generation, and sample utilization. This perspective enables us to identify key differences among existing methods and clarify the overall design space. Building on this framework, we develop a comprehensive and fair evaluation protocol, benchmarking representative DiffDA methods across diverse low-data classification tasks. Extensive experiments reveal the relative strengths and limitations of different DiffDA strategies and offer practical insights into method design and deployment. All methods are re-implemented within a unified codebase, with full release of code and configurations to ensure reproducibility and to facilitate future research.