FDIP-NET: a physics-inspired deepfake image detector leveraging FDIP luminance dynamics and calibrated CNNs
Frontiers in Imaging · 30.10.2026 · 5
DOI: 10.3389/fimag.2026.1935077 ↗Kurz gefasst
Fachartikel in der Zeitschrift Frontiers in Imaging, online erschienen am 30.10.2026. Verfasst von Saman Maghsoudi und 2 weiteren Autor:innen. Die Arbeit nennt 58 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
Introduction
Synthetic faces generated using generative models are becoming increasingly indistinguishable from real ones, making reliable detection important for trust, security, and authenticity in digital media.
Methods
We propose FDIP-NET, a physics-inspired framework based on Fluid-Dynamics Inspired Processing (FDIP) that fuses RGB appearance with three luminance-gradient descriptors: a divergence-like response, gradient magnitude, and a kinetic-energy-like response. The framework combines robustness-oriented training, temperature scaling, and risk-aware threshold selection. Experiments were conducted on three publicly available facial-image datasets, DFRI, DF-wild, and RFF-140K, with an additional evaluation on FaceForensics++ c23 using its official train, validation, and test splits.
Results
FDIP-NET achieved mean held-out test accuracies of 98.654%, 98.952%, and 99.669% on DFRI, DF-wild, and RFF-140K, respectively, and 91.00% frame-level accuracy on FaceForensics++ c23.
Discussion
Component-wise comparisons indicate that luminance-gradient descriptors can complement RGB features, with their contribution depending on the dataset and selected representation. Strong within-dataset performance does not establish reliable transfer to unseen dataset distributions.
Zitieren
APA
Maghsoudi, S., Sangeeta, S., & Kanwal, N. (2026). FDIP-NET: a physics-inspired deepfake image detector leveraging FDIP luminance dynamics and calibrated CNNs. Frontiers in Imaging, 5. https://doi.org/10.3389/fimag.2026.1935077
Vancouver
Maghsoudi S, Sangeeta S, Kanwal N. FDIP-NET: a physics-inspired deepfake image detector leveraging FDIP luminance dynamics and calibrated CNNs. Frontiers in Imaging. 2026;5. doi:10.3389/fimag.2026.1935077
BibTeX
@article{Maghsoudi2026,
author = {Maghsoudi, Saman and Sangeeta, Sangeeta and Kanwal, Nadia},
title = {FDIP-NET: a physics-inspired deepfake image detector leveraging FDIP luminance dynamics and calibrated CNNs},
journal = {Frontiers in Imaging},
year = {2026},
volume = {5},
doi = {10.3389/fimag.2026.1935077}
}
EUVIMED