Research map / Federated learning and privacy / Personalized and heterogeneous FL
FedDifRC: Unlocking the Potential of Text-to-Image Diffusion Models in Heterogeneous Federated Learning
Huan Wang, Haoran Li, Huaming Chen, Jun Yan, Jiahua Shi, Jun Shen
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 |
|---|---|---|---|
| add TDCL (ablation) | baseline · CIFAR10 (NID1/NID2 average) | accuracy: 80.41 -> 82.81 | Improved |
| add NDCR (ablation) | baseline · CIFAR10 (NID1/NID2 average) | accuracy: 80.41 -> 82.40 | Improved |
| add NDCR to TDCL (ablation) | TDCL · CIFAR10 (NID1/NID2 average) | accuracy: 82.81 -> 84.29 | Improved |
| add TDCL (ablation) | baseline · CIFAR10-LT (NID1_0.2, rho average) | accuracy: 60.46 -> 61.31 | Improved |
| add NDCR (ablation) | baseline · CIFAR10-LT (NID1_0.2, rho average) | accuracy: 60.46 -> 61.50 | Improved |
| add NDCR to TDCL (ablation) | TDCL · CIFAR10-LT (NID1_0.2, rho average) | accuracy: 61.31 -> 62.25 | Improved |
| use fused multi-layer diffusion features instead of best single layer (L=2) (ablation) | FedDifRC · CIFAR10 NID1_0.05 | accuracy: 82.60 -> 83.14 | Improved |
| use fused multi-layer diffusion features instead of best single layer (L=3) (ablation) | FedDifRC · CIFAR10 NID1_0.2 | accuracy: 87.81 -> 88.27 | Improved |
| use fused multi-layer diffusion features instead of best single layer (L=2) (ablation) | FedDifRC · CIFAR10 NID1_0.5 | accuracy: 89.03 -> 89.31 | Improved |
| use fused multi-layer diffusion features instead of best single layer (L=2) (ablation) | FedDifRC · CIFAR10 NID2 | accuracy: 75.73 -> 76.45 | Improved |
| use only layer L=1 features instead of fused features (ablation) | FedDifRC · CIFAR10 NID1_0.05 | accuracy: 83.14 -> 80.21 | Hurt |
| align raw Stable Diffusion features with L2 loss added to FedRCL instead of FedDifRC design | FedRCL · CIFAR10 NID1_0.05 | accuracy: 82.30 -> 80.95 | Hurt |
Closest papers by meaning
- Exploring One-Shot Semi-supervised Federated Learning with Pre-trained Diffusion Models (AAAI 2024 · similarity 0.92)
- One-Shot Heterogeneous Federated Learning with Local Model-Guided Diffusion Models (ICML 2025 · similarity 0.89)
- Confusion-Resistant Federated Learning via Diffusion-Based Data Harmonization on Non-IID Data (NeurIPS 2024 · similarity 0.88)
- Progressive Distribution Matching for Federated Semi-Supervised Learning (AAAI 2025 · similarity 0.88)
- FedBiP: Heterogeneous One-Shot Federated Learning with Personalized Latent Diffusion Models (CVPR 2025 · similarity 0.87)
- FedQS: Optimizing Gradient and Model Aggregation for Semi-Asynchronous Federated Learning (NeurIPS 2025 · similarity 0.87)
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.
