Next Article in Journal
Gait Classification Based on Micro-Doppler Effect
Previous Article in Journal
Unified Modulation Matrix-Based Shared Control for Teleoperated Multi-Robot Formation and Obstacle Avoidance
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Sensors
Volume 26
Issue 8
10.3390/s26082389
sensors-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

DIFC-Net: Diffusion-Intrinsic Feature Capture for AI-Generated Image Detection

by
Shaofeng Lu
1,
Jin Tian
1,*,
Yujin Zhang
1,
Fei Wu
1,
Li Gong
2 and
Han Pan
3
1
School of Electric and Electronic Engineering, Shanghai University of Engineering Science, Shanghai 201620, China
2
School of Geographic Sciences, East China Normal University, Shanghai 200241, China
3
School of Aeronautics and Astronautics, Shanghai Jiao Tong University, Shanghai 200240, China
*
Author to whom correspondence should be addressed.
Sensors 2026, 26(8), 2389; https://doi.org/10.3390/s26082389
Submission received: 26 February 2026 / Revised: 3 April 2026 / Accepted: 7 April 2026 / Published: 13 April 2026
(This article belongs to the Section Sensing and Imaging)
Browse Figures
Versions Notes

Abstract

Diffusion models (e.g., Stable Diffusion, DALL·E 3) can now generate images that are nearly indistinguishable from real ones, making synthetic image detection increasingly challenging. We propose DIFC-Net, a diffusion-intrinsic detection framework that identifies AI-generated images by analyzing their reconstruction behavior during diffusion inversion rather than relying on visual artifacts. DIFC-Net jointly captures spatial discrepancy signals and latent diffusion trajectory evolution, and adaptively fuses them into a unified forensic representation. Extensive cross-model evaluations show that DIFC-Net achieves 90.29% average AUC on multiple unseen diffusion generators, outperforming state-of-the-art detectors while maintaining strong generalization without relying on training-time knowledge of specific generative models.
Keywords: diffusion model detection; image forensic analysis; latent diffusion inversion; residual discrepancy; multimodal feature fusion; synthetic image authenticity diffusion model detection; image forensic analysis; latent diffusion inversion; residual discrepancy; multimodal feature fusion; synthetic image authenticity

Share and Cite

MDPI and ACS Style

Lu, S.; Tian, J.; Zhang, Y.; Wu, F.; Gong, L.; Pan, H. DIFC-Net: Diffusion-Intrinsic Feature Capture for AI-Generated Image Detection. Sensors 2026, 26, 2389. https://doi.org/10.3390/s26082389

AMA Style

Lu S, Tian J, Zhang Y, Wu F, Gong L, Pan H. DIFC-Net: Diffusion-Intrinsic Feature Capture for AI-Generated Image Detection. Sensors. 2026; 26(8):2389. https://doi.org/10.3390/s26082389

Chicago/Turabian Style

Lu, Shaofeng, Jin Tian, Yujin Zhang, Fei Wu, Li Gong, and Han Pan. 2026. "DIFC-Net: Diffusion-Intrinsic Feature Capture for AI-Generated Image Detection" Sensors 26, no. 8: 2389. https://doi.org/10.3390/s26082389

APA Style

Lu, S., Tian, J., Zhang, Y., Wu, F., Gong, L., & Pan, H. (2026). DIFC-Net: Diffusion-Intrinsic Feature Capture for AI-Generated Image Detection. Sensors, 26(8), 2389. https://doi.org/10.3390/s26082389

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop