AI Paper Atlas

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

Learning Phase Mask for Privacy-Preserving Passive Depth Estimation

Zaid Tasneem, Giovanni Milione, Yi-Hsuan Tsai, Xiang Yu, Ashok Veeraraghavan, Manmohan Chandraker, Francesco Pittaluga

ECCV 2022 · Federated learning and privacy › Privacy attacks and secure learning · approach: Visual privacy and face de-identification · 17 citations (OpenAlex)

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
add heavy defocus (no phase mask) to the sensorall-in-focus sensor · VGGFace2 face identificationface identification accuracy (%): 80.1% -> 21.7%Improved
add privacy adversary to phase mask optimization (sensor D vs depth-only sensor C) (ablation)depth-optimized phase mask · VGGFace2 face identificationface identification accuracy (%): 40% -> 20.5%Improved
use a random phase mask instead of no mask (ablation)sensor A · depth estimation and face identificationprivacy-utility trade-offHurt
use a larger undiffracted PSF (focal distance u=0.1m vs 0.17 and 1.0m) (ablation)learned phase mask sensor · privacy-utility trade-offprivacy leakageImproved
reduce sensor resolution from 64x64 to 16x16 (ablation)learned phase mask sensor · VGGFace2 face identificationface identification accuracyImproved
reduce sensor resolution from 64x64 to 16x16 (ablation)learned phase mask sensor · NYUv2 depthdepth estimation qualityNo clear effect
replace all-in-focus sensor with learned sensor Hall-in-focus sensor · VGGFace2 face identificationface identification accuracy (%): 80.1% -> 5.3%Improved
replace conventional sensor with learned sensor for face verification attackconventional sensor · LFW face verificationAUC: 0.99 -> 0.67Improved
replace conventional camera with prototype sensorconventional camera · Real face verification datasetAUC: 0.89 -> 0.59Improved
use learned sensor outputs instead of Kinect depth for action recognitionI3D on Kinect · NTURGBD120 cross-setuptop-1 accuracy (%): 79.1% -> 70.5%Hurt
move from simulation to real prototype sensorlearned sensor H · NYUv2 / real capturesdelta1 accuracy (%): 84.69% -> 83.73%No clear effect

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