UGAD: Universal Generative AI Detector utilizing Frequency Fingerprints

TL;DR

This paper presents UGAD, a multi-modal AI-generated image detector that uses frequency fingerprints and radial features to significantly improve accuracy over existing methods.

Contribution

The study introduces UGAD, a novel multi-modal detection framework combining frequency analysis and deep learning for robust AI-generated image detection.

Findings

12.64% increase in detection accuracy

28.43% increase in AUC over state-of-the-art methods

Effective differentiation of real and AI-generated images

Abstract

In the wake of a fabricated explosion image at the Pentagon, an ability to discern real images from fake counterparts has never been more critical. Our study introduces a novel multi-modal approach to detect AI-generated images amidst the proliferation of new generation methods such as Diffusion models. Our method, UGAD, encompasses three key detection steps: First, we transform the RGB images into YCbCr channels and apply an Integral Radial Operation to emphasize salient radial features. Secondly, the Spatial Fourier Extraction operation is used for a spatial shift, utilizing a pre-trained deep learning network for optimal feature extraction. Finally, the deep neural network classification stage processes the data through dense layers using softmax for classification. Our approach significantly enhances the accuracy of differentiating between real and AI-generated images, as evidenced by a 12.64% increase in accuracy and 28.43% increase in AUC compared to existing state-of-the-art methods.