AI Paper Atlas

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

FED-PsyAU: Privacy-Preserving Micro-Expression Recognition via Psychological AU Coordination and Dynamic Facial Motion Modeling

Jingting Li, Yu Qian, Lin Zhao, Su-Jing Wang

ICCV 2025 · Federated learning and privacy › Personalized and heterogeneous FL · approach: Federated medical and multimodal learning · 4 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 AU features (LFE) with dual-stream InceptionNet (ablation)full-face OF baseline · DFME (ablation)UF1: 0.3249 -> 0.3362Improved
add spatial structure encoder (SSE) over ROIs (ablation)FED-PsyAU LFE · DFME (ablation)UF1: 0.3362 -> 0.3480Improved
add local (upper/lower) DPK-GAT topology modeling (ablation)FED-PsyAU without GAT · DFME (ablation)UF1: 0.3480 -> 0.3629Improved
add global DPK-GAT topology modeling (ablation)FED-PsyAU without GAT · DFME (ablation)UF1: 0.3480 -> 0.3604Improved
combine local and global GAT (ablation)FED-PsyAU · DFME (ablation)UF1: 0.3629 -> 0.3853Improved
group ROIs by AU for feature extraction (ablation)FED-PsyAU · DFME (ablation)UF1: 0.3683 -> 0.3853Improved
add psychological prior adjacency (ablation)DPK-GAT · DFME (ablation)UF1: 0.3615 -> 0.3723Improved
add data-driven prior attention (ablation)DPK-GAT · DFME (ablation)UF1: 0.3615 -> 0.3705Improved
combine psychological and data-driven priors (ablation)DPK-GAT · DFME (ablation)UF1: 0.3615 -> 0.3853Improved
add psychological prior on top of data-driven prior (ablation)DPK-GAT · DFME (ablation)UF1: 0.3705 -> 0.3853Improved
use FED-PsyAU MER network instead of FRDFME · DFME (7-class, 10-fold CV)UF1: 0.3559 -> 0.3853Improved
use FED-PsyAU instead of HTNetCAS(ME)^3 · CAS(ME)^3 (3-class, LOSO)UF1: 0.5767 -> 0.6221Improved

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