AI Paper Atlas

Research map / Federated learning and privacy / Privacy attacks and secure learning

Vulnerability of Privacy-Preserving Visual Localization against Diffusion-based Attacks

Maxime Pietrantoni, Torsten Sattler, Gabriela Csurka

ECCV 2026 · Federated learning and privacy › Privacy attacks and secure learning · approach: Visual privacy and face de-identification

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
invert with diffusion model instead of FFConv feed-forward network (high sparsity)inversion attack · Cambridge Landmarks, SIFT 500 kptsLPIPS: 0.44 -> 0.28Improved
invert with diffusion model instead of FFConv (medium sparsity)inversion attack · Cambridge Landmarks, SIFT 1K kptsFID: 198 -> 97Improved
invert with diffusion model instead of FFConv (low sparsity)inversion attack · Cambridge Landmarks, SIFT 2K kptsFID: 248 -> 130Improved
invert with diffusion model instead of FFUnetSA (SegLoc)inversion attack · Cambridge LandmarksFID: 189 -> 91Improved
invert with diffusion model instead of FFConv (RandLine SIFT)inversion attack · Cambridge LandmarksFID: 158 -> 73Improved
invert with diffusion model instead of FFTrans (CoordPerm SIFT)inversion attack · Cambridge LandmarksFID: 220 -> 76Improved
increase keypoint density from 500 to 2K per image (ablation)FFConv inversion · Cambridge LandmarksSSIM: 0.45 -> 0.38Hurt
increase keypoint density from 500 to 2K per image (ablation)diffusion inversion · Cambridge LandmarksSSIM: 0.49 -> 0.41Hurt
condition on DPT segmentations instead of SuperPoint descriptorsinversion attack · 7-ScenesFID: 96.8 -> 222.5Improved
condition on DPT segmentations instead of SuperPoint descriptorsinversion attack · 7-ScenesRD(SSIM): 0.71 -> 0.26Improved
use finer segmentation granularity (SAM2 instead of DPT)inversion attack · 7-ScenesLPIPS: 0.59 -> 0.50Hurt
use SegLoc segmentation instead of RandLine (SIFT) for localizationPPVL method · Cambridge Landmarksmedian translation error (m): 0.07 -> 0.22Hurt

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy