InfoResearchPreprint
SmoothShift: Disrupting Visual Adversarial Attacks on LVLMs at Test Time
- Published
- Record updated
Summary
SmoothShift is a lightweight, training-free defense against visual adversarial attacks on large vision-language models (LVLMs). It combines local smoothing, which weakens fine-grained perturbation patterns, with spatial transformations that disrupt their alignment with the encoder's patch grid. The authors report improved robustness across three LVLM families, four datasets and eight attack methods, with clean performance largely preserved.
Mitigation
SmoothShift: a training-free defense combining local smoothing and spatial transformations. It requires no model updates, auxiliary networks, or additional model inference.
Topics
Related items
- MediumGemini AI agents repeatedly bypassed safeguards to attack real websiteSimilar attack · BleepingComputer
- InfoSTA: A Split Temporal Adversary Method for Evaluating Robustness of Multimodal RL-based Driving PoliciesSimilar attack · Arxiv (cs.RO + cs.CV security)
- InfoCracks in Embeddings: Robustness and Stability of JEPA-Based Chest X-Ray RepresentationsSimilar attack · Arxiv (cs.RO + cs.CV security)
- LowAdversarial Tracking: Attacks on Image Sensors for Object Tracking in Autonomous VehiclesSimilar attack · Arxiv (cs.RO + cs.CV security)
- InfoQuantum-inspired orthogonal adapters for certifiably robust medical image classificationSimilar attack · Arxiv (cs.RO + cs.CV security)