Research map / Federated learning and privacy / Personalized and heterogeneous FL
Fed-CO$_{2}$: Cooperation of Online and Offline Models for Severe Data Heterogeneity in Federated Learning
Zhongyi Cai, Ye Shi, Wei Huang, Jingya Wang
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 changed | Where | Measured | Effect |
|---|---|---|---|
| use online-offline prediction fusion with personalized online Batch Normalization | Fed-CO2 · Office-Caltech10 feature skew | average accuracy: 70.5±2.0 -> 74.5±0.3 | Improved |
| use online-offline prediction fusion with personalized online Batch Normalization | Fed-CO2 · DomainNet feature skew | average accuracy: 48.0±1.0 -> 51.8±0.2 | Improved |
| use online-offline prediction fusion | Fed-CO2 · CIFAR10 Pathological | test accuracy: 89.05±0.04 -> 88.79±0.25 | Mixed |
| use online-offline prediction fusion | Fed-CO2 · CIFAR10 Dirichlet | test accuracy: 75.41±0.37 -> 77.45±0.30 | Improved |
| use online-offline prediction fusion | Fed-CO2 · CIFAR100 Pathological | test accuracy: 54.96±1.30 -> 58.50±0.43 | Improved |
| use online-offline prediction fusion | Fed-CO2 · CIFAR100 Dirichlet | test accuracy: 28.70±0.46 -> 32.43±0.37 | Improved |
| use online-offline prediction fusion | Fed-CO2 · Digits with label skew and feature skew | average accuracy: 89.70±0.06 -> 92.59±0.15 | Improved |
| add intra-client knowledge transfer (ablation) | Fed-CO2 · DomainNet feature skew | average accuracy: 51.12±0.19 -> 51.83±0.25 | Improved |
| add inter-client knowledge transfer (ablation) | Fed-CO2 · DomainNet feature skew | average accuracy: 49.27±0.46 -> 51.83±0.25 | Improved |
| add intra-client and inter-client knowledge transfer (ablation) | Fed-CO2 · DomainNet feature skew | average accuracy: 47.88±0.70 -> 51.83±0.25 | Improved |
Closest papers by meaning
- Exploring Vacant Classes in Label-Skewed Federated Learning (AAAI 2025 · similarity 0.91)
- Exploiting Label Skews in Federated Learning with Model Concatenation (AAAI 2024 · similarity 0.90)
- An Aggregation-Free Federated Learning for Tackling Data Heterogeneity (CVPR 2024 · similarity 0.90)
- FedGPS: Statistical Rectification Against Data Heterogeneity in Federated Learning (NeurIPS 2025 · similarity 0.89)
- A Simple Data Augmentation for Feature Distribution Skewed Federated Learning (CVPR 2025 · similarity 0.89)
- FedWon: Triumphing Multi-domain Federated Learning Without Normalization (ICLR 2024 · similarity 0.89)
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.
