Problem
Research questions and friction points this paper is trying to address.
adversarial examples
multi-view
3D reconstruction
VGGT
perturbation consistency
Generating Multi-view Adversarial Examples for Visual Geometry Grounded Transformer
🤖 AI Summary
本文针对VGGT模型的安全漏洞，提出MVAP-G方法生成多视角一致的对抗扰动，无需逐场景优化即可有效攻击3D重建系统。
📝 Abstract
The Visual Geometry Grounded Transformer (VGGT) enables unified feed-forward 3D reconstruction from multi-view images. However, deploying such a high-performance model may expose critical security vulnerabilities. Traditional adversarial perturbations require costly per-scene optimization, while Universal Adversarial Perturbations (UAPs) rely on a single static pattern and fail to effectively attack VGGT. To address these limitations, we propose \textbf{MVAP-G}, a multi-view adversarial perturbation generator that produces imperceptible consistent perturbations across multiple views in a single feed-forward pass. To ensure perturbation consistency across diverse scenes, we design a cross-view adversarial alignment mechanism to process multi-view images. Experiments demonstrate that MVAP-G significantly degrades VGGT performance without iterative optimization during inference. This work pioneers multi-view adversarial attacks on 3D foundation models, uncovering severe vulnerabilities and underscoring the urgent need for robust 3D vision systems. The code is available at https://github.com/qsong2001/mvap-g.
Innovation
Methods, ideas, or system contributions that make the work stand out.
MVAP-G
multi-view adversarial perturbation
cross-view adversarial alignment
single feed-forward pass
🔎 Similar Papers
No similar papers found.
