Research map / Federated learning and privacy / Personalized and heterogeneous FL
FLAIR: Federated Learning Annotated Image Repository
Congzheng Song, Filip Granqvist, Kunal Talwar
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 |
|---|---|---|---|
| federated instead of centralized training, coarse labels, random init | ResNet-18 · FLAIR coarse | per-class AP: 60.40 -> 50.41 | Hurt |
| federated instead of centralized training, coarse labels, ImageNet init | ResNet-18 · FLAIR coarse | per-class AP: 67.71 -> 62.09 | Hurt |
| add central differential privacy to federated training, coarse, random init | ResNet-18 · FLAIR coarse | per-class AP: 50.41 -> 28.80 | Hurt |
| add central differential privacy to federated training, coarse, ImageNet init | ResNet-18 · FLAIR coarse | per-class AP: 62.09 -> 44.28 | Hurt |
| ImageNet pretrained init instead of random init, centralized coarse | ResNet-18 · FLAIR coarse | per-class AP: 60.40 -> 67.71 | Improved |
| ImageNet pretrained init instead of random init, private federated coarse | ResNet-18 · FLAIR coarse | per-class AP: 28.80 -> 44.28 | Improved |
| federated instead of centralized training, fine labels, random init | ResNet-18 · FLAIR fine | per-class AP: 14.90 -> 1.53 | Hurt |
| federated instead of centralized training, fine labels, ImageNet init | ResNet-18 · FLAIR fine | per-class AP: 20.26 -> 2.03 | Hurt |
| add central differential privacy, fine labels, ImageNet init | ResNet-18 · FLAIR fine | overall AP: 27.31 -> 12.67 | Hurt |
| ImageNet pretrained init instead of random init, federated fine | ResNet-18 · FLAIR fine | overall AP: 22.68 -> 27.31 | Improved |
| federated training instead of centralized on binary structure label | ResNet-18 · FLAIR binary structure | AP: 87.72 -> 84.22 | Hurt |
| add differential privacy on binary structure label, ImageNet init | ResNet-18 · FLAIR binary structure | AP: 90.49 -> 83.41 | Hurt |
Closest papers by meaning
- TurboSVM-FL: Boosting Federated Learning through SVM Aggregation for Lazy Clients (AAAI 2024 · similarity 0.85)
- BSemiFL: Semi-supervised Federated Learning via a Bayesian Approach (ICML 2025 · similarity 0.85)
- pFedES: Generalized Proxy Feature Extractor Sharing for Model Heterogeneous Personalized Federated Learning (AAAI 2025 · similarity 0.85)
- Language-Guided Transformer for Federated Multi-Label Classification (AAAI 2024 · similarity 0.85)
- Exploring One-Shot Semi-supervised Federated Learning with Pre-trained Diffusion Models (AAAI 2024 · similarity 0.85)
- FedWon: Triumphing Multi-domain Federated Learning Without Normalization (ICLR 2024 · similarity 0.84)
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.
