AI Paper Atlas

Research map / Federated learning and privacy / Personalized and heterogeneous FL

Benchmarking Federated Learning & Knowledge Distillation for Point Cloud Classification

Aizierjiang Aiersilan

ECCV 2026 · Federated learning and privacy › Personalized and heterogeneous FL · approach: Federated medical and multimodal learning

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
train with federated label-skew non-IID instead of centralizedPointNet++ SSG (best FL method FedNova) · ModelNet40instance accuracy: 92.26 -> 76.32Hurt
train with federated label-skew non-IID instead of centralizedPointNet++ SSG (best FL method FedProx) · Craniosynostosisinstance accuracy: 100.00 -> 75.83Hurt
use server-side optimizers (FedAvgM, FedAdam, FedYogi, FedAdagrad) instead of FedAvgPointNet++ SSG · ModelNet40instance accuracy: 58.51 -> 4.05Hurt
use FedProx proximal term instead of FedAvgPointNet++ SSG · ModelNet40instance accuracy: 58.51 -> 71.04Improved
use MOON model-contrastive term instead of FedAvgPointNet++ SSG · ModelNet40instance accuracy: 58.51 -> 45.84Hurt
use FedNova instead of FedProxPointNet++ SSG · Craniosynostosisinstance accuracy: 75.83 -> 50.83Hurt
distill teacher into compact student (74.51% smaller)SmallPointNet2 · ModelNet40instance accuracy: 92.44 -> 92.74Improved
use Logit-MSE instead of Vanilla KDSmallPointNet2 · ModelNet40instance accuracy: 92.61 -> 87.07Hurt
keep hard-label cross-entropy term in KD (Logit-MSE) with collapsed SCAFFOLD teacher (ablation)SmallPointNet2 · ModelNet40 combinedinstance accuracy: 8.50 -> 92.94Improved
remove hard-label cross-entropy (feature/attention objectives) with collapsed teacher (ablation)SmallPointNet2 · ModelNet40 combinedinstance accuracy: 92.94 -> 16.99Hurt
use chance-level FedAvgM teacher with hard-label Vanilla KD (ablation)SmallPointNet2 · Craniosynostosis combinedinstance accuracy: 99.6 -> 97.5No clear effect
use chance-level teacher with pure-transfer Attention Transfer (ablation)SmallPointNet2 · Craniosynostosis combinedinstance accuracy: 83.3 -> 25.0Hurt

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