AI Paper Atlas

Research map / Architectures and efficiency / Attention and transformers

AUFormer: Vision Transformers are Parameter-Efficient Facial Action Unit Detectors

Kaishen Yuan, Zitong Yu, Xin Liu, Weicheng Xie, Huanjing Yue, Jingyu Yang

ECCV 2024 · Architectures and efficiency › Attention and transformers · approach: Token merging and masked pretraining · 39 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
replace full fine-tuning with PETL (Convpass) (ablation)ViT-B/16 AU detector · BP4Daverage F1: 61.9 -> 63.6Improved
add collaboration mechanism over a single Convpass (ablation)AUFormer · BP4Daverage F1: 63.6 -> 64.3Improved
add MRF and CA operators in MoKE (ablation)AUFormer · BP4Daverage F1: 64.3 -> 65.1Improved
replace WA-Loss with MDWA-Loss (gamma and m) (ablation)full fine-tuned ViT baseline · BP4Daverage F1: 61.9 -> 63.4Improved
add truncation margin m in MDWA-Loss (ablation)AUFormer · BP4Daverage F1: 65.6 -> 66.2Improved
add difficulty-aware gamma in MDWA-Loss (ablation)AUFormer · BP4Daverage F1: 65.1 -> 65.6Improved
use Convpass instead of Adapter, LoRA or VPT as adaptation module (ablation)PETL AU detector · BP4Daverage F1: 63.1 / 63.0 / 62.6 -> 63.6Improved
use dilated rates 1, 3, 5 instead of rate 1 only (ablation)MRF operator · BP4Daverage F1: 65.6 -> 66.2Improved
add dilation rate 2 (1,2,3 vs 1,3) (ablation)MRF operator · BP4Daverage F1: 65.9 -> 65.9No clear effect
enlarge CA neighborhood from 3x3 to 5x5 or 7x7 (ablation)CA operator · BP4Daverage F1: 66.2 -> 65.9 (5x5), 66.0 (7x7)Hurt
increase truncation margin m from 0.1 to 0.2 (ablation)MDWA-Loss · BP4Daverage F1: 66.2 -> 65.4Hurt
apply AUFormer in BP4D to DISFA transfer vs FG-NetAUFormer · BP4D to DISFAaverage F1: 54.4 (FG-Net) vs 43.6Hurt

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