AI Paper Atlas

Research map / Federated learning and privacy / Differential privacy

Privacy-Preserving Visual Learning Using Doubly Permuted Homomorphic Encryption

Ryo Yonetani, Vishnu Naresh Boddeti, Kris M. Kitani, Yoichi Sato

ICCV 2017 · Federated learning and privacy › Differential privacy · approach: Private online and user-level learning · 73 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
increase sparsity of local classifier updates from 0.01% to 95.6% (ablation)DPHE · Caltech101accuracy (%): 89.7 -> 88.2No clear effect
increase sparsity of local classifier updates from 0.01% to 95.6% (ablation)DPHE · Caltech101encryption time (sec): 620 -> 62Improved
increase sparsity from 0.01% to 79.1% (ablation)DPHE · Caltech101encryption time (sec): 620 -> 186Improved
increase number of users N from 5 to 100 with fixed, partially duplicated images per user (ablation)DPHE · Caltech101accuracy (%)No clear effect
increase number of users N from 5 to 100 with non-duplicated images split among users (ablation)DPHE · Caltech101accuracy (%)Hurt
use DPHE (no added noise, iterative averaging) instead of RA12privacy-preserving classifier learning · Caltech101accuracy (%): 83.8 -> 89.3Improved
use DPHE instead of PRR10privacy-preserving classifier learning · Caltech101accuracy (%): 41.6 -> 89.3Improved
use DPHE instead of RA12privacy-preserving classifier learning · Caltech256accuracy (%): 68.0 -> 74.7Improved
use DPHE instead of RA12privacy-preserving classifier learning · CelebA (40 attributes)mean accuracy (%): 64 -> 84Improved
use DPHE instead of PRR10privacy-preserving classifier learning · CelebA (40 attributes)mean accuracy (%): 78 -> 84Improved
use DPHE instead of RA12privacy-preserving classifier learning · Life-logging sensitive place detectionaverage precision: 0.704 -> 0.729Improved
add privacy preservation via DPHE relative to non-private No-PP SVMlinear SVM · YouTube8M subset (top 100 attributes)mAP: 0.48 -> 0.47No 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