Research map / Graph learning / Knowledge and temporal graphs
Upper-body Hierarchical Graph for Skeleton Based Emotion Recognition in Assistive Driving
Jiehui Wu, Jiansheng Chen, Qifeng Luo, Siqi Liu, Youze Xue, Huimin Ma
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 changed | Where | Measured | Effect |
|---|---|---|---|
| add fully connected edges to the physically connected hierarchical graph (ablation) | UbH-GCN · AIDE | Acc. (%): 74.55 -> 75.37 | Improved |
| replace conventional CTR-GCN graph with UbH hierarchical graph (PC edges) (ablation) | CTR-GCN · AIDE | Acc. (%): 71.92 -> 74.55 | Improved |
| use imbalance-aware logit variation loss instead of cross-entropy (ablation) | UbH-GCN 4-ensemble · AIDE | Acc. (%): 74.55 -> 77.50 | Improved |
| apply imbalance-aware loss to HD-GCN (ablation) | HD-GCN 4-ensemble · AIDE | Acc. (%): 76.03 -> 76.68 | Improved |
| use four-way ensemble (two roots x joint/bone) instead of two-way (ablation) | UbH-GCN · AIDE | Acc. (%): 75.37 -> 77.50 | Improved |
| use upper-body skeleton instead of full-body input | UbH-GCN · Emilya | Acc. (%): 93.01 -> 91.28 (3-fold), 94.42 -> 93.98 (10-fold) | Mixed |
| drop motion streams from ensemble | UbH-GCN · AIDE | Acc. (%) | No clear effect |
Closest papers by meaning
- Take an Emotion Walk: Perceiving Emotions from Gaits Using Hierarchical Attention Pooling and Affective Mapping (ECCV 2020 · similarity 0.85)
- ARBEE: Towards Automated Recognition of Bodily Expression of Emotion in the Wild (IJCV 2019 · similarity 0.85)
- MMTL-UniAD: A Unified Framework for Multimodal and Multi-Task Learning in Assistive Driving Perception (CVPR 2025 · similarity 0.84)
- Understanding Emotional Body Expressions via Large Language Models (AAAI 2025 · similarity 0.84)
- GaitCycFormer: Leveraging Gait Cycles and Transformers for Gait Emotion Recognition. (AAAI 2025 · similarity 0.83)
- Facial-R1: Aligning Reasoning and Recognition for Facial Emotion Analysis (AAAI 2026 · similarity 0.82)
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.
