Generalized Synthetic Image Detection with Enhanced RGB-Noise Representation Learning

2026-07-07

AI Summary

This work addresses the limited robustness of existing image forgery detection methods under cross-generator generalization and real-world image degradation scenarios. To this end, the authors propose RNSIDNet, a dual-branch framework that jointly learns RGB semantic and high-frequency noise features. The semantic branch employs an attention-enhanced CLIP backbone to capture global semantics, while the noise branch leverages FiLM-modulated Bayar convolutions to dynamically adapt noise representations. Additionally, Hard Sample-aware Contrastive Learning (HSCL) is introduced to refine feature space distribution. Evaluated on eight public benchmarks, RNSIDNet achieves state-of-the-art performance, demonstrating significantly improved generalization, robustness against unseen generative models and complex degradations, and enhanced computational efficiency.

Abstract

The rapid advancement of large-scale generative models has accelerated the spread of highly deceptive AI-generated images, making generalized synthetic image detection a critical imperative. Existing forensic networks often struggle with cross-model generalization and realworld degradations due to their reliance on single-domain representations and conventional binary classification optimization. To overcome these limitations, we propose RNSIDNet, a novel forensic framework that achieves robust detection through enhanced RGB-Noise representation learning. Specifically, our method employs a dual-branch architecture where global RGB semantics, extracted by an attention-refined CLIP backbone, dynamically modulate highfrequency noise artifacts captured by Bayar convolutions via a Feature-wise Linear Modulation (FiLM) module. To further enhance the learned representations, we design a Hard Sample-aware Contrastive Learning (HSCL) strategy. By explicitly penalizing challenging training samples, HSCL reshapes the latent feature space to maximize the discriminative margin between pristine and synthetic domains. Extensive experiments across eight public benchmark datasets verify that our model achieves state-of-the-art performance, delivering superior generalization ability, robustness, and computational efficiency. Code and dataset will be publicly available on https://github.com/multimediaFor/RNSIDNet.

Problem

• synthetic image detection
• cross-model generalization
• real-world degradations
• generalized detection
• AI-generated images

Innovation

• RGB-Noise Representation Learning
• Feature-wise Linear Modulation (FiLM)
• Hard Sample-aware Contrastive Learning (HSCL)
• Generalized Synthetic Image Detection
• Dual-branch Architecture