GenMix: Effective Data Augmentation with Generative Diffusion Model Image Editing

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Islam, Khawar, Zaheer, Muhammad Zaigham, Mahmood, Arif, Nandakumar, Karthik, Akhtar, Naveed
Format: Preprint
Veröffentlicht: 2024
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866912613963661312
author Islam, Khawar
Zaheer, Muhammad Zaigham
Mahmood, Arif
Nandakumar, Karthik
Akhtar, Naveed
author_facet Islam, Khawar
Zaheer, Muhammad Zaigham
Mahmood, Arif
Nandakumar, Karthik
Akhtar, Naveed
contents 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.
format Preprint
id arxiv_https___arxiv_org_abs_2412_02366
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle GenMix: Effective Data Augmentation with Generative Diffusion Model Image Editing
Islam, Khawar
Zaheer, Muhammad Zaigham
Mahmood, Arif
Nandakumar, Karthik
Akhtar, Naveed
Computer Vision and Pattern Recognition
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.
title GenMix: Effective Data Augmentation with Generative Diffusion Model Image Editing
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2412.02366