AI Paper Atlas

Research map / Federated learning and privacy / Communication-efficient FL

Towards Privacy-preserved Pre-training of Remote Sensing Foundation Models with Federated Mutual-guidance Learning

Jieyi Tan, Chengwei Zhang, Bo Dang, Yansheng Li

ICCV 2025 · Federated learning and privacy › Communication-efficient FL · approach: Efficient local training in FL · 5 citations (OpenAlex)

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 changedWhereMeasuredEffect
add Server-to-Clients Guidance (SCG) to SST (ablation)FedSense · RESISC-45OA: 94.52 -> 94.70Improved
add SCG to SST (ablation)FedSense · LoveDAmIoU: 51.87 -> 51.85No clear effect
add Clients-to-Server Guidance (CSG) to SST (ablation)FedSense · RESISC-45OA: 94.52 -> 95.25Improved
add CSG to SST (ablation)FedSense · DIOR-RmAP: 64.83 -> 65.12Improved
add both SCG and CSG to SST (ablation)FedSense · RESISC-45OA: 94.52 -> 96.33Improved
add both SCG and CSG to SST (ablation)FedSense · LEVIR-CD+F1: 72.43 -> 73.21Improved
use proposed quantization instead of uniform quantization (1-bit) (ablation)FedSense · RESISC-45OA: 95.56 -> 96.01Improved
use proposed quantization instead of FedPAQ (1-bit) (ablation)FedSense · RESISC-45OA: 95.79 -> 96.01Improved
1-bit quantized communication vs full precision (ablation)FedSense · Downstream tasksaverage performance: minimal degradation (0.23% across tasks)No clear effect
increase number of clients from 2 to 10 (80K to 800K samples) (ablation)FedSense · RESISC-45OA: 94.50 -> 96.33Improved
increase number of clients from 2 to 10 (80K to 800K samples) (ablation)FedSense · LoveDAmIoU: 51.56 -> 52.74Improved
scale backbone from Swin-T to Swin-L (ablation)FedSense · LoveDAmIoU: 52.74 -> 55.96Improved

Closest papers by meaning

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.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy