AI Paper Atlas

Research map / Federated learning and privacy / Communication-efficient FL

SpecGradFilter: A Spectral Gradient Filtering Framework for Taming Federated Heterogeneity

Liyang Yuan, Yibo Yang, Dandan Guo, Peter Richtárik, Zhouchen Lin

TPAMI 2026 · Federated learning and privacy › Communication-efficient FL · approach: Client participation analysis · 0 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
use low-frequency suppression instead of high-frequency suppression (ablation)SpecGradFilter · CIFAR-10 (alpha=0.1)accuracy: 59.99 -> 78.67Improved
use low-frequency suppression instead of random suppression (ablation)SpecGradFilter · CIFAR-100 (alpha=0.1)accuracy: 28.74 -> 40.87Improved
use spectral filtering instead of l2-norm rescaling (ablation)SpecGradFilter · CIFAR-100 (alpha=0.1)accuracy: 27.04 -> 40.87Improved
use FFT low-frequency suppression versus spatial Gaussian detrend (ablation)SpecGradFilter · CIFAR-100 (alpha=0.1)accuracy: 41.53 vs 40.87No clear effect
filter all convolutional layers instead of deep stage only (ablation)SpecGradFilter · CIFAR-10 (alpha=0.1)accuracy: 73.89 -> 78.67Improved
also filter classifier head (ablation)SpecGradFilter · CIFAR-10 (alpha=0.1)accuracy: 78.67 -> 78.58No clear effect
filter only patch embedding of ViT instead of all eligible weights (ablation)SpecGradFilter · CIFAR-10 ViT (alpha=0.1)accuracy: 44.77 -> 45.16Improved
filter attention in ViT (ablation)SpecGradFilter · CIFAR-100 ViT (alpha=0.1)accuracy: 23.11 -> 22.50Hurt
use natural flatten order 0123 instead of 3210 (ablation)SpecGradFilter · CIFAR-10 (alpha=0.1)accuracy: 55.87 -> 78.05Improved
use 1D FFT instead of 2D FFT (ablation)SpecGradFilter · CIFAR-100 (alpha=0.1)accuracy: 31.30 -> 40.16Improved
apply filter in centralized IID trainingSpecGradFilter · CIFAR-100 centralizedaccuracy: 68.66 -> 67.09Hurt
apply filter in near-IID federated learning alpha=100SpecGradFilter vs FedAvg · CIFAR-10accuracy: 81.28 -> 86.74Improved

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