Time Step Generating: A Universal Synthesized Deepfake Image Detector

TL;DR

This paper introduces a universal deepfake image detector called Time Step Generating (TSG) that leverages diffusion model features and time step control to distinguish real from synthetic images with high accuracy and generalizability.

Contribution

The proposed TSG method does not depend on pre-trained model reconstruction or specific datasets, offering a more robust and universal approach to synthetic image detection.

Findings

Achieves significant accuracy improvements on GenImage benchmark.

Demonstrates high generalizability across different datasets.

Does not rely on pre-trained generative model reconstruction.

Abstract

Currently, high-fidelity text-to-image models are developed in an accelerating pace. Among them, Diffusion Models have led to a remarkable improvement in the quality of image generation, making it vary challenging to distinguish between real and synthesized images. It simultaneously raises serious concerns regarding privacy and security. Some methods are proposed to distinguish the diffusion model generated images through reconstructing. However, the inversion and denoising processes are time-consuming and heavily reliant on the pre-trained generative model. Consequently, if the pre-trained generative model meet the problem of out-of-domain, the detection performance declines. To address this issue, we propose a universal synthetic image detector Time Step Generating (TSG), which does not rely on pre-trained models' reconstructing ability, specific datasets, or sampling algorithms. Our method utilizes a pre-trained diffusion model's network as a feature extractor to capture fine-grained details, focusing on the subtle differences between real and synthetic images. By controlling the time step t of the network input, we can effectively extract these distinguishing detail features. Then, those features can be passed through a classifier (i.e. Resnet), which efficiently detects whether an image is synthetic or real. We test the proposed TSG on the large-scale GenImage benchmark and it achieves significant improvements in both accuracy and generalizability.