← Back to all news
arXiv Computer Vision By Anton Nuzhdin, Marcel Worring, Ivona Najdenkoska

FUSED: Forensic-Semantic Mixture-of-Experts for AI Inpainting Detection and Localization

Read the original on arXiv Computer Vision →

FUSED is a new framework that jointly detects and localizes AI-generated inpainting by combining low-level forensic cues with high-level semantic features through a sparsely-gated Mixture-of-Experts architecture. It predicts both an image-level manipulation score and a pixel-level mask of the inpainted region. On the OpenSDID cross-generator benchmark, FUSED outperforms existing methods, especially on unseen generators, and transfers effectively to the AutoSplice and CocoGlide benchmarks, doubling localization performance.

Machine-generated by The Flow from the publisher's headline and feed description — not written or checked by a human. The full article lives at arXiv Computer Vision.

Related stories

arXiv Computer Vision
Sep 3

From Detection to Localization: A Unified Forensics Framework for Fully Synthetic and Tampered Images

The paper proposes a unified forensics framework that extends traditional binary image manipulation detection to a multiclass setting—distinguishing real, fully synthetic, and tampered images. It adds a segmentation branch for pixel‑level localization of tampered regions, achieving higher classification accuracy and IoU scores compared to recent benchmarks. The authors provide the implementation on GitHub for reproducibility.

By Annalisa Gallina, Marco Fiorucci, Marco Brigo, Federica Battisti, Lamberto Ballan
computer-visionroboticsbenchmarks
More like this →
arXiv Computer Vision
Sep 30

RED: Reconstruction Evolution Dynamics for Generalizable AI-Generated Image Detection

The paper introduces RED (Reconstruction Evolution Dynamics), a new framework for detecting AI-generated images that leverages the evolution of intermediate reconstruction stages rather than relying solely on static representations or endpoint discrepancies. RED uses a frozen multiscale VQ‑VAE and a frozen CLIP encoder to capture a reconstruction trajectory, then learns image‑adaptive stage weights from token negative log‑likelihoods provided by a frozen VAR model. Experiments on six benchmarks show RED achieves the highest average accuracy (92.5%) and precision (97.5%) among evaluated methods, and it remains robust to common image degradations.

By Wenpeng Mu, Junshan Jin, Tanfeng Sun, Xinghao Jiang, Qiang Xu
benchmarks
More like this →