AI Paper Atlas

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

Gradient Inversion with Generative Image Prior

Jinwoo Jeon, Jaechang Kim, Kangwook Lee, Sewoong Oh, Jungseul Ok

NeurIPS 2021 · Federated learning and privacy › Privacy attacks and secure learning · approach: Gradient inversion attacks · 3 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 generative prior search (GI-x to GI-z/w)GI-x · ImageNet 64x64 val, ResNet18, B=4PSNR: 13.78 -> 15.58Improved
add generative prior search (GI-x+BN to GI-z/w+BN)GI-x+BN · ImageNet 64x64 val, ResNet18, B=4PSNR: 15.52 -> 16.31Improved
add generative prior search (GI-x to GI-z/w)GI-x · ImageNet 64x64 val, ResNet18, B=4LPIPS: 0.4376 -> 0.3023Improved
add parameter space search after latent search (GI-z to GI-z/w) (ablation)GIAS · ImageNet 64x64 val, ResNet18, B=4PSNR: 14.27 -> 15.58Improved
add latent space search before parameter search (GI-w to GI-z/w) (ablation)GIAS · ImageNet 64x64 val, ResNet18, B=4PSNR: 14.70 -> 15.58Improved
optimize only parameters of untrained generator instead of pixels (GI-x to GI-w) (ablation)GI-x · ImageNet 64x64 val, ResNet18, B=4PSNR: 13.78 -> 14.70Improved
optimize only parameters of untrained generator instead of pixels (GI-x to GI-w) (ablation)GI-x · ImageNet 64x64 val, ResNet18, B=4LPIPS: 0.4376 -> 0.5121Hurt
GIAS without BN versus GI-x with BN statisticsGIAS · ImageNet 64x64 val, ResNet18, B=4PSNR: 15.58 vs 15.52No clear effect
use latent search then parameter search instead of latent then input pixels (GI-z/x) (ablation)GI-z · ImageNet 64x64 val, ResNet18, B=4visual qualityImproved
add Gaussian gradient noise 0.01GIAS · ImageNet 64x64 val, ResNet18, B=4PSNRNo clear effect
add large Gaussian gradient noise 0.1GIAS · ImageNet 64x64 val, ResNet18, B=4PSNRHurt
increase batch size to 32GIAS · ImageNet 64x64 val, ResNet18, B=4PSNRHurt

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