Fractal Characterization of Low-Correlation Signals in AI-Generated Image Detection

Authors: Wenwei Xie, Jie Yin, Lu Ma, Xuansong Zhang, Wenjing Zhang

Published: 2026-04-19 05:55:34+00:00

Comment: https://github.com/jim-xie-cn/Research-Deepfake

AI Summary

This paper introduces a novel signal-level approach for detecting AI-generated images by identifying intrinsic discrepancies between synthetic and authentic content. It demonstrates that low-correlation signals serve as distinctive markers for deepfakes, proposing a fractal theory-based method to quantify these signals and capture subtle statistical anomalies. The approach shows superior detection performance and emphasizes a new research direction for deepfake detection.

Abstract

AI-generated imagery has reached near-photorealistic fidelity, yet this technology poses significant threats to information security and societal trust. Existing deepfake detection methods often exhibit limited robustness in open-world scenarios. To address this limitation, this paper investigates intrinsic discrepancies between synthetic and authentic images from a signal-level perspective. Our analysis reveals that low-correlation signals serve as distinctive markers for differentiating AI-generated imagery from real photographs. Building on this insight, we introduce a novel method for quantifying these signals based on fractal theory. By analyzing the fractal characteristics of low-correlation signals, our method effectively captures the subtle statistical anomalies inherent to the synthesis process. Extensive experimental results demonstrate the method's robustness and superior detection performance. This work emphasizes the need to shift research focus to a new signal-level direction for deepfake detection. Theoretically, this proposed approach is not limited to face image identification but can be applied to all AI-generated image detection tasks. This study provides a new research direction for deepfake detection.

Key findings

The study found that low-correlation signals possess high discriminative power for differentiating AI-generated images from real ones, especially after suppressing dominant high-correlation signals via PCA. Fractal-based features (Fractal Dimension, Information Entropy, and Multifractal Spectrum) extracted from residual images showed significantly enhanced distinctions, with high Kolmogorov–Smirnov (KS) statistics consistently exceeding 0.93 for real vs. fake images.

Approach

The method first uses Principal Component Analysis (PCA) to decompose images, separating high-correlation (main object) signals from low-correlation (residual) signals. It then applies fractal theory, calculating features like fractal dimension, multifractal spectrum, lacunarity, and information entropy from these residual images to quantify and distinguish subtle statistical anomalies characteristic of AI-generated content.

Datasets

1-million-fake-faces, FFHQ

Model(s)

Principal Component Analysis (PCA), Fractal Dimension (FD), Multifractal Spectrum (MFS), Lacunarity, Information Entropy

Author countries

China

← Previous