CloakDiff: Reversible Adversarial Examples for Privacy Protection in Vision-Language Models

Jul 14, 2026

Researchers introduce CloakDiff, a novel framework that generates imperceptible and reversible adversarial examples to protect privacy against text-based query attacks on Vision-Language Models (VLMs). CloakDiff uniquely combines diffusion-based adversarial editing with an invertible network, enabling lossless recovery of the original image while maintaining high visual quality and strong cross-model transferability. Experimental results show effective multimodal privacy preservation across multiple datasets and VLMs.

Why it matters: This work presents a significant advance in privacy protection for VLMs, allowing users to safeguard sensitive image attributes without compromising image quality or recoverability.

Full story at: arXiv Computer Vision ↗