Unlearnable examples keep publicly shared photos from being learned by unauthorized face-recognition models. An imperceptible perturbation, added before sharing, makes any model trained on the protected photos fail on clean faces. The perturbation is crafted on the shared image, however the attacker trains on the face it extracts, cropped and resized to the recognizer input, and under this extraction the protection collapses. We propose LPID, which builds the extraction into the unlearnable-example objective. LPID confines the perturbation to the extracted face region and optimizes it through a differentiable model of the extraction, concentrating its energy in the frequency band the extraction preserves. Because this robustness is a property of the transform rather than of any identity, LPID is re-optimized per album and protects even users it has never seen. LPID attains the lowest attacker accuracy of all methods in every setting we evaluate, holding the attacker below $10\%$ under crop+resize extraction on identities unseen at protection time, while remaining imperceptible at $32.7$\,dB PSNR and $0.161$ LPIPS.
Large-scale and diverse datasets are needed to train AI models to take real-time decisions for autonomous vehicles (AVs), an intelligent transportation system (ITS) application. Pedestrian intention and trajectory prediction are critical models used in AVs, requiring datasets inv…
Table extraction from business documents relies on a cascaded pipeline where Table Detection (TD) first localizes tables and Table Structure Recognition (TSR) then recovers their internal layout. Building task-specific training sets for this pipeline is costly, particularly for T…
Face-swapping deepfakes pose an escalating threat to personal privacy by enabling unauthorized identity manipulation. While adversarial approaches have demonstrated success against black-box face recognition (FR) models, their applicability to face-swapping scenarios remains unde…
Vision Language Models (VLMs) offer powerful multimodal ability but also expose users to text-based privacy attacks where adversaries crawl online photos and query VLMs to extract sensitive attributes. Existing reversible adversarial example (RAE) methods protect images in purely…
Split face recognition reduces client-side computation but exposes intermediate features to feature inversion attacks and unauthorized analysis by honest-but-curious (HBC) servers. Existing privacy-preserving face recognition methods mainly aim to resist unauthorized reconstructi…
The rise of customized diffusion models has fueled a boom in personalized visual content creation, but it also introduces serious risks of malicious misuse, thereby posing threats to personal privacy. Image aesthetics are strongly correlated with human perception of image quality…