A Multi-Domain Feature Fusion Framework for Generalizable Deepfake Detection Across Different Generators
Abstract
Abstract Deepfakes are artificially generated images, audio, or videos that threaten privacy, security, and information integrity. Detecting such content has become increasingly challenging as recent generative models produce highly realistic synthetic media with substantially different generation characteristics. Existing deepfake image detectors commonly rely on spatial, gradient, or frequency-domain artifacts in isolation, making their learned representations sensitive to generator-specific distributions and limiting their transferability to unseen generative models. This paper proposes the Spatial-Gradient-Frequency Fusion Network (SGFF-Net), a multi-domain deepfake image detection framework that integrates spatial, gradient, and Discrete Wavelet Transform (DWT)-based frequency representations within a dual residual learning architecture. Rather than evaluating detection only under conventional intra-dataset settings, SGFF-Net is systematically assessed under cross-model, cross-paradigm, multi-source, and real-world distribution shifts. The results indicate that multi-domain feature fusion provides more transferable forensic representations than individual spatial, gradient, or frequency-based representations, although the degree of transfer depends on the source and target generator distributions. Cross-paradigm experiments further highlight an asymmetric transfer between GAN- and diffusion-based generators, demonstrating that robustness across one generator family does…
Read the paper
Related studies
- Learning from First-Generation Qualitative Approaches in the IS Discipline: An Evolutionary View and Some Implications for Authors and Evaluators (PART 1/2)
- Commentary on the clinical and preclinical dosage limits of interstitially administered magnetic fluids for therapeutic hyperthermia based on current practice and efficacy models
- Prosthesis: New Technological Opportunities and Innovative Biomedical Devices
- The history and future of neural modeling for cochlear implants
- Ryanodine receptors: waking up from refractoriness
- Novel Transgenic Rabbit Model Sheds Light on the Puzzling Role of Matrix Metalloproteinase-12 in Atherosclerosis
Go deeper in Evidence Hub
Evidence Hub adds a study data sheet (design, sample size, evidence level, key numbers), an AI summary and a chat grounded in the paper's text.