Research map / Federated learning and privacy / Personalized and heterogeneous FL
Federated CLIP for Resource-Efficient Heterogeneous Medical Image Classification
Yihang Wu, Ahmad Chaddad
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 local masked MLP classifier with ensemble prediction (ablation) | FedMedCLIP · ISIC2019 | AVG accuracy (%): 71.94 -> 78.44 | Improved |
| add class-wise KL regularization L_sim (ablation) | FedMedCLIP · ISIC2019 | AVG accuracy (%): 78.44 -> 80.4 | Improved |
| add class-wise KL regularization L_sim (ablation) | FedMedCLIP · ISIC2019 client C1 | accuracy (%): -1.21% on C1 | Hurt |
| add federated FAM aggregation (ablation) | FedMedCLIP · ISIC2019 | AVG accuracy (%): 70.41 -> 80.4 | Improved |
| add float16 + zlib model compression (ablation) | FedMedCLIP · ISIC2019 | AVG accuracy (%): 80.48 -> 80.4 | No clear effect |
| use larger softmax scale T=10 instead of T=2 (ablation) | FedMedCLIP · ISIC2019 | AVG accuracy (%): 80.4 -> 78.94 | Hurt |
| use ViT-L/14 backbone instead of ViT-B/32 (ablation) | FedMedCLIP · ISIC2019 | AVG accuracy (%): 80.4 -> 81.83 | Improved |
| use FedMedCLIP instead of FedAPT | FedMedCLIP vs FedAPT · BraTS global test | accuracy (%): 74.04 -> 64.66 | Hurt |
| use FedMedCLIP instead of FedAVG/FedAPT | FedMedCLIP vs baselines · ISIC2019 global client | accuracy (%): 85.43 -> 81.01 | Hurt |
| use FedMedCLIP instead of second-best baseline | FedMedCLIP vs FedAPT · ISIC2019 AVG | accuracy (%): 72.21 -> 80.4 | Improved |
| use FedMedCLIP under FGSM attack | FedMedCLIP · ISIC2019 | AVG accuracy (%): 33.85 (vs ~10% drop for FedCLIP/LoRA) | Improved |
Closest papers by meaning
- CLIP-Guided Federated Learning on Heterogeneity and Long-Tailed Data (AAAI 2024 · similarity 0.89)
- Overcoming Heterogeneous Data in Federated Medical Vision-Language Pre-training: A Triple-Embedding Model Selector Approach (AAAI 2025 · similarity 0.88)
- FedVCK: Non-IID Robust and Communication-Efficient Federated Learning via Valuable Condensed Knowledge for Medical Image Analysis (AAAI 2025 · similarity 0.87)
- FedPIA – Permuting and Integrating Adapters Leveraging Wasserstein Barycenters for Finetuning Foundation Models in Multi-Modal Federated Learning (AAAI 2025 · similarity 0.87)
- F^3OCUS - Federated Finetuning of Vision-Language Foundation Models with Optimal Client Layer Updating Strategy via Multi-objective Meta-Heuristics (CVPR 2025 · similarity 0.86)
- FedBiP: Heterogeneous One-Shot Federated Learning with Personalized Latent Diffusion Models (CVPR 2025 · similarity 0.85)
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.
