AI Paper Atlas

Research map / Federated learning and privacy / Personalized and heterogeneous FL

Federated Learning with Local Openset Noisy Labels

Zonglin Di, Zhaowei Zhu, Xiaoxiao Li, Yang Liu

ECCV 2024 · Federated learning and privacy › Personalized and heterogeneous FL · approach: Federated semi-supervised and graph learning · 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 globally shared DP contrastive labels (FedDPCont) to FedAvgFedAvg · CIFAR-10 symmetric noise 0.2accuracy: 76.84 -> 84.77Improved
add globally shared DP contrastive labels (FedDPCont) to FedAvgFedAvg · CIFAR-10 random noise 0.4accuracy: 59.19 -> 72.69Improved
apply centralized noisy-label methods (loss correction, Co-teaching, T-revision) locally in FLFedAvg · CIFAR-10accuracy: 76.84 -> 79.14 (LC, symmetric 0.2)No clear effect
vary label DP level epsilon from 0.01 to 100 versus no DP (ablation)FedDPCont · CIFAR-10 random noise 0.4accuracy: 74.03 (No DP) vs 72.44 (epsilon 0.81)No clear effect
replace DivideMix with FedDPCont on CIFAR-10 symmetric 0.2DivideMix · CIFAR-10 symmetric noise 0.2accuracy: 79.28 -> 84.77Improved
use FedDPCont instead of DivideMix on CIFAR-10 symmetricDivideMix · CIFAR-10 symmetric noisetraining time (hr): 22.33 -> 4.17Improved
use FedDPCont instead of DivideMix at high noise on CIFAR-100DivideMix · CIFAR-100 symmetric noise 0.8accuracy: 19.87 -> 14.04Hurt
use FedDPCont instead of FedCorr on CIFAR-100FedCorr · CIFAR-100 symmetric noise 0.2accuracy: 57.65 -> 58.55Improved
use FedDPCont on real-world Clothing-1M instead of FedAvgFedAvg · Clothing-1Maccuracy: 70.27 -> 70.88Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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