AI Paper Atlas

Research map / Federated learning and privacy / Differential privacy

Decouple-and-Sample: Protecting Sensitive Information in Task Agnostic Data Release

Abhishek Singh, Ethan Garza, Ayush Chopra, Praneeth Vepakomma, Vivek Sharma, Ramesh Raskar

ECCV 2022 · Federated learning and privacy › Differential privacy · approach: Private data analysis frameworks · 7 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
replace Suppression with DP-Sampling in local sampling (ablation)Sanitizer · E5 UTKFaceutility (CAS): 0.208 -> 0.521Improved
replace Suppression with DP-Sampling in local sampling (ablation)Sanitizer · E5 CelebAutility (CAS): 0.7042 -> 0.817Improved
replace DP-Obfuscation (Laplace noise) with DP-Sampling (ablation)Sanitizer · E5 UTKFaceutility (CAS): 0.208 -> 0.521Improved
replace DP-Obfuscation with DP-Sampling (ablation)Sanitizer · E5 CelebAleakage: 0.7129 -> 0.6955Improved
Sanitizer vs best baseline on privacy-utility AuCSanitizer · UTKFace-GAuC: 0.439 -> 0.487Improved
Sanitizer vs TIPRDC on privacy-utility AuCSanitizer · Fairface-RAuC: 0.441 -> 0.476Improved
Sanitizer vs Adversarial Noise on CAS utilitySanitizer · E4 CelebAutility (CAS): 0.746 -> 0.9022Improved
Sanitizer vs Adversarial Noise on CAS leakageSanitizer · E4 CelebAleakage: 0.663 -> 0.6955Hurt
Sanitizer vs Adversarial Noise on CAS leakageSanitizer · E4 UTKFaceleakage: 0.695 -> 0.474Improved
remove each global-decoupler component (aligner, distance correlation, adversary) (ablation)global-decoupler · UTKFaceleakage/utility trade-offImproved
use global-decoupler instead of beta-VAE with latent interpolation (ablation)Sanitizer · UTKFace (E2 setup)privacy-utility trade-offImproved

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