How much of an image does a detector need to see? Small RGB regions can retain useful evidence of image synthesis even when they reveal little of the full scene. Motivated by single-patch detection, we study patch-wise supervision: a shared backbone classifies explicit crops, each crop receives its own loss, and patch...
Zhi-Da Zhang, Tao Wu, Si-Yu Liu et al.· 0 citations
The Band-Attention Modulation Network (BAM-Net) is proposed, a novel framework that pioneers learnable, fine-grained modulation of frequency components for forgery detection and exhibits exceptional generalization in cross-dataset, cross-compression, and cross-manipulation scenarios.
Zhi-Da Zhang, Wen-Kui Yang, Xin Ma et al.· 1 citation
EIB-Net is proposed, an Entropy-guided Information Bottleneck Network that introduces a novel Image Entropy metric to automatically select the most informative patch, then processes it with a Variational Information Bottleneck (VIB) to learn compact, generalizable features.
While recent advances in 3D generation have enabled impressive visual synthesis, existing methods often rely on 2D diffusion supervision without explicit mechanisms for geometric consistency, leading to spatial hallucinations such as duplicated structures and misaligned geometry. These issues become more severe in 4D g...
Hong-Bo Wang, Huaibo Huang, Jie Cao et al.· 0 citations
Energy-Guided Flow Matching is introduced that explicitly models a coarse-to-fine generative trajectory by moving endpoint that evolves smoothly from low-frequency image to clean image and requires no adaptation of the backbone and training data.
Haoyang Tong, Yu He, Fang Li et al.· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
