Research map / Federated learning and privacy / Personalized and heterogeneous FL
No One Left Behind: Real-World Federated Class-Incremental Learning
Jiahua Dong, Hongliu Li, Yang Cong, Gan Sun, Yulun Zhang, Luc Van Gool
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 category-balanced gradient-adaptive compensation loss (ablation) | LGA · CIFAR-100 (T=5) | avg accuracy: 66.7 -> 73.0 | Improved |
| add category-balanced gradient-adaptive compensation loss (ablation) | LGA · MiniImageNet (T=20) | avg accuracy: 60.1 -> 62.9 | Improved |
| add semantic distillation loss (ablation) | LGA · CIFAR-100 (T=5) | avg accuracy: 60.4 -> 73.0 | Improved |
| add semantic distillation loss (ablation) | LGA · MiniImageNet (T=10) | avg accuracy: 49.2 -> 67.5 | Improved |
| add semantic distillation loss (ablation) | LGA · CIFAR-100 (T=20) | avg accuracy: 56.4 -> 70.6 | Improved |
| add semantic distillation loss (ablation) | LGA · TinyImageNet (T=10) | avg accuracy: 45.3 -> 53.2 | Improved |
| add proxy server reconstruction (ablation) | LGA · CIFAR-100 (T=5) | avg accuracy: 71.9 -> 73.0 | Improved |
| add proxy server reconstruction (ablation) | LGA · TinyImageNet (T=5) | avg accuracy: 45.7 -> 54.5 | Improved |
| add proxy server reconstruction (ablation) | LGA · ImageNet-1000 (T=5) | avg accuracy: 60.1 -> 63.5 | Improved |
| add proxy server reconstruction (ablation) | LGA · MiniImageNet (T=20) | avg accuracy: 61.1 -> 62.9 | Improved |
| add proxy server reconstruction (ablation) | LGA · CIFAR-100 (T=20) | avg accuracy: 67.1 -> 70.6 | Improved |
| add category-balanced gradient-adaptive compensation loss (ablation) | LGA · MiniImageNet (T=10) | avg accuracy: 61.0 -> 67.5 | Improved |
Closest papers by meaning
- Federated Class-Incremental Learning: A Hybrid Approach Using Latent Exemplars and Data-Free Techniques to Address Local and Global Forgetting (ICLR 2025 · similarity 0.92)
- Federated Orthogonal Training: Mitigating Global Catastrophic Forgetting in Continual Federated Learning (ICLR 2024 · similarity 0.90)
- Accurate Forgetting for Heterogeneous Federated Continual Learning (ICLR 2024 · similarity 0.89)
- Towards Efficient Replay in Federated Incremental Learning (CVPR 2024 · similarity 0.87)
- C$^2$Prompt: Class-aware Client Knowledge Interaction for Federated Continual Learning (NeurIPS 2025 · similarity 0.87)
- Class-wise Balancing Data Replay for Federated Class-Incremental 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.
