AI Paper Atlas

Research map / Label-efficient and robust learning / Recognition and long-tail

From Facial Expression Recognition to Interpersonal Relation Prediction

Zhanpeng Zhang, Ping Luo, Chen Change Loy, Xiaoou Tang

IJCV 2017 · Label-efficient and robust learning › Recognition and long-tail · approach: Uncertainty-aware facial expression recognition · 314 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 attribute propagation (ablation)DCN · AFLW+CelebA+ExpW averagebalanced accuracy: 76.1 -> 80.9Improved
add attribute propagation (ablation)DCN · goateebalanced accuracy: 63.3 -> 73.5Improved
add attribute propagation (ablation)DCN · narrow eyesbalanced accuracy: 94.2 -> 94.5No clear effect
add attribute propagation (ablation)DCN · youngbalanced accuracy: 84.4 -> 83.5Hurt
add attribute propagation (ablation)DCN (no fine-tuning) · SFEW 2.0accuracy: 45.51% -> 49.77%Improved
add attribute propagation (ablation)DCN (fine-tuned) · SFEW 2.0accuracy: 52.06% -> 55.27%Improved
increase alternating iterations from M=1 to M=10 (ablation)DCN+AP · AFLW+CelebA+ExpW averagebalanced accuracy: 78.4 -> 80.9Improved
pre-train the DCN on expression attributes instead of random init (ablation)S-DCN · relation prediction (movie subset)balanced accuracy: 62.42 -> 65.36Improved
pre-train the DCN on age attributes (ablation)S-DCN · relation prediction (movie subset)balanced accuracy: 62.42 -> 64.67Improved
pre-train the DCN on gender attributes (ablation)S-DCN · relation prediction (movie subset)balanced accuracy: 62.42 -> 63.10Improved
pre-train the DCN on pose attributes (ablation)S-DCN · relation prediction (movie subset)balanced accuracy: 62.42 -> 62.83Improved
pre-train on all attributes (ablation)S-DCN · relation prediction (movie subset)balanced accuracy: 62.42 -> 68.17Improved

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