Research map / 3D vision / Human pose estimation
Take an Emotion Walk: Perceiving Emotions from Gaits Using Hierarchical Attention Pooling and Affective Mapping
Uttaran Bhattacharya, Christian Roncal, Trisha Mittal, Rohan Chandra, Kyra Kapsaskis, Kurt Gray, Aniket Bera, Dinesh Manocha
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 affective loss (labeled only) (ablation) | baseline · Emotion-Gait (test) | mAP: 0.67 -> 0.76 | Improved |
| add hierarchical pooling (labeled only) (ablation) | baseline · Emotion-Gait (test) | mAP: 0.67 -> 0.72 | Improved |
| add both AL and HP (labeled only) (ablation) | baseline · Emotion-Gait (test) | mAP: 0.67 -> 0.78 | Improved |
| add unlabeled data with autoencoder (no AL or HP) (ablation) | baseline · Emotion-Gait (test) | mAP: 0.67 -> 0.70 | Improved |
| add affective loss with all data (ablation) | Hierarchical pooling model · Emotion-Gait (test) | mAP: 0.78 -> 0.84 | Improved |
| add hierarchical pooling with all data (ablation) | Affective loss model · Emotion-Gait (test) | mAP: 0.81 -> 0.84 | Improved |
| add autoencoder and unlabeled data (full model) (ablation) | labeled-only AL+HP · Emotion-Gait (test) | mAP: 0.78 -> 0.84 | Improved |
| add unlabeled data on angry class (ablation) | baseline no AL no HP · Emotion-Gait (test) | AP angry: 0.51 -> 0.55 | Improved |
| add unlabeled data on neutral class (ablation) | baseline no AL no HP · Emotion-Gait (test) | AP neutral: 0.42 -> 0.48 | Improved |
| replace STEP with proposed method | STEP · Emotion-Gait (test) | mAP: 0.77 -> 0.84 | Improved |
| add more unlabeled data | Our Method · Emotion-Gait (test) | mAP | Improved |
| vary lambda_quat and lambda_aff between 0.5 and 2.5 | Our Method · Emotion-Gait (test) | performance | No clear effect |
Closest papers by meaning
- GaitCycFormer: Leveraging Gait Cycles and Transformers for Gait Emotion Recognition. (AAAI 2025 · similarity 0.87)
- Understanding Emotional Body Expressions via Large Language Models (AAAI 2025 · similarity 0.85)
- $E^3$: Exploring Embodied Emotion Through A Large-Scale Egocentric Video Dataset (NeurIPS 2024 · similarity 0.84)
- FEEL: Quantifying Heterogeneity in Physiological Signals for Generalizable Emotion Recognition (NeurIPS 2025 · similarity 0.84)
- AffectGPT: A New Dataset, Model, and Benchmark for Emotion Understanding with Multimodal Large Language Models (ICML 2025 · similarity 0.84)
- H-MoRe: Learning Human-centric Motion Representation for Action Analysis (CVPR 2025 · similarity 0.83)
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.
