# Take an Emotion Walk: Perceiving Emotions from Gaits Using Hierarchical Attention Pooling and Affective Mapping

arXiv: https://arxiv.org/abs/1911.08708 · HTML: https://mlsanity.com/paper/1911.08708/

## Key facts

- Cited by 10 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 20 Nov 2019.

## Built on by (most cited first, 10 of 10)

- [Speech2AffectiveGestures: Synthesizing Co-Speech Gestures with Generative Adversarial Affective Expression Learning](https://mlsanity.com/paper/2108.00262/) (2021)
- [ChallenCap: Monocular 3D Capture of Challenging Human Performances using Multi-Modal References](https://mlsanity.com/paper/2103.06747/) (2021)
- [Unlocking the Emotional World of Visual Media: An Overview of the Science, Research, and Impact of Understanding Emotion](https://mlsanity.com/paper/2307.13463/) (2023)
- [PsyMo: A Dataset for Estimating Self-Reported Psychological Traits from Gait](https://mlsanity.com/paper/2308.10631/) (2023)
- [TNTC: two-stream network with transformer-based complementarity for gait-based emotion recognition](https://mlsanity.com/paper/2110.13708/) (2021)
- [MSA-GCN:Multiscale Adaptive Graph Convolution Network for Gait Emotion Recognition](https://mlsanity.com/paper/2209.08988/) (2022)
- [Emotion Recognition from Skeleton Data: A Comprehensive Survey](https://mlsanity.com/paper/2507.18026/) (2025)
- [Speech2UnifiedExpressions: Synchronous Synthesis of Co-Speech Affective Face and Body Expressions from Affordable Inputs](https://mlsanity.com/paper/2406.18068/) (2024)
- [CGTGait: Collaborative Graph and Transformer for Gait Emotion Recognition](https://mlsanity.com/paper/2509.16623/) (2025)
- [Single-Stream Multi-Feature Fusion with Temporal Robustness for Gait Emotion Recognition](https://mlsanity.com/paper/2609.11680/) (2026)

## Abstract (arXiv)

We present an autoencoder-based semi-supervised approach to classify perceived human emotions from walking styles obtained from videos or motion-captured data and represented as sequences of 3D poses. Given the motion on each joint in the pose at each time step extracted from 3D pose sequences, we hierarchically pool these joint motions in a bottom-up manner in the encoder, following the kinematic chains in the human body. We also constrain the latent embeddings of the encoder to contain the space of psychologically-motivated affective features underlying the gaits. We train the decoder to reconstruct the motions per joint per time step in a top-down manner from the latent embeddings. For the annotated data, we also train a classifier to map the latent embeddings to emotion labels. Our semi-supervised approach achieves a mean average precision of 0.84 on the Emotion-Gait benchmark dataset, which contains both labeled and unlabeled gaits collected from multiple sources. We outperform current state-of-art algorithms for both emotion recognition and action recognition from 3D gaits by 7%--23% on the absolute. More importantly, we improve the average precision by 10%--50% on the absolute on classes that each makes up less than 25% of the labeled part of the Emotion-Gait benchmark dataset.