GenMix: Effective Data Augmentation with Generative Diffusion Model Image Editing

TL;DR

GenMix introduces a prompt-guided generative data augmentation method using image editing and fractal patterns to improve visual classification performance across various domains and tasks.

Contribution

This paper presents GenMix, a novel data augmentation approach that enhances in-domain and cross-domain classification through prompt-guided image editing and blending techniques.

Findings

Improves classification accuracy on eight public datasets.

Enhances adversarial robustness and performance in data-scarce scenarios.

Outperforms existing state-of-the-art augmentation methods.

Abstract

Data augmentation is widely used to enhance generalization in visual classification tasks. However, traditional methods struggle when source and target domains differ, as in domain adaptation, due to their inability to address domain gaps. This paper introduces GenMix, a generalizable prompt-guided generative data augmentation approach that enhances both in-domain and cross-domain image classification. Our technique leverages image editing to generate augmented images based on custom conditional prompts, designed specifically for each problem type. By blending portions of the input image with its edited generative counterpart and incorporating fractal patterns, our approach mitigates unrealistic images and label ambiguity, improving the performance and adversarial robustness of the resulting models. Efficacy of our method is established with extensive experiments on eight public datasets for general and fine-grained classification, in both in-domain and cross-domain settings. Additionally, we demonstrate performance improvements for self-supervised learning, learning with data scarcity, and adversarial robustness. As compared to the existing state-of-the-art methods, our technique achieves stronger performance across the board.