Search research articles

ABOUT JoVE

Overview Leadership Blog JoVE Help Center

AUTHORS

Publishing Process Editorial Board Scope & Policies Peer Review FAQ Submit

LIBRARIANS

Testimonials Subscriptions Access Resources Library Advisory Board FAQ

RESEARCH

JoVE Journal Methods Collections JoVE Encyclopedia of Experiments Archive

EDUCATION

JoVE Core JoVE Business JoVE Science Education JoVE Lab Manual Faculty Resource Center Faculty Site

Terms & Conditions of Use

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by

Same author

Ontology-based annotation and fuzzy recommendation for community formation in smart city knowledge platforms.

Frontiers in artificial intelligence·2026

Same author

Fast and scalable annotation-free LV-centered ROI localisation in stress perfusion cardiac MRI.

Computers in biology and medicine·2026

Same author

Real-world road damage dataset with potholes, cracks, and maintenance holes.

Scientific reports·2026

Same author

Spatio-temporal graph autoencoder for automated evaluation of human actions in 3D in immersive VR-based training for archaeologists.

Scientific reports·2026

Same author

Data-related Ablation for Reinforcing Deep Learning in Explaining Complex Phenomena.

International journal of neural systems·2026

Same author

Correction: A Benchmark Dataset for Radio Signal Image-based Person Re-Identification.

Scientific data·2025

Search research articles

Related Experiment Video

Updated: Sep 3, 2025

Estimation of Contact Regions Between Hands and Objects During Human Multi-Digit Grasping

Estimation of Contact Regions Between Hands and Objects During Human Multi-Digit Grasping

Published on: April 21, 2023

Affective Action and Interaction Recognition by Multi-View Representation Learning from Handcrafted Low-Level

Danilo Avola¹, Marco Cascio¹, Luigi Cinque¹

¹Department of Computer Science, Sapienza University of Rome, Via Salaria 113, Rome 00198, Italy.

International Journal of Neural Systems

|July 26, 2022

Summary

This study introduces a novel method for recognizing human emotions from body movements in videos. The approach uses skeletal data and achieves high accuracy in identifying affective actions and interactions.

Keywords:

Affective action affective interaction bag-of-visual-words handcrafted low-level skeleton features multi-view representation learning

More Related Videos

Capturing Dynamic Finger Gesturing with High-resolution Surface Electromyography and Computer Vision

Capturing Dynamic Finger Gesturing with High-resolution Surface Electromyography and Computer Vision

Published on: March 28, 2025

Author Spotlight: Addressing Technical and Subjective Challenges in Measuring Classroom Attention

Author Spotlight: Addressing Technical and Subjective Challenges in Measuring Classroom Attention

Published on: December 15, 2023

Related Experiment Videos

Last Updated: Sep 3, 2025

Estimation of Contact Regions Between Hands and Objects During Human Multi-Digit Grasping

Estimation of Contact Regions Between Hands and Objects During Human Multi-Digit Grasping

Published on: April 21, 2023

Capturing Dynamic Finger Gesturing with High-resolution Surface Electromyography and Computer Vision

Capturing Dynamic Finger Gesturing with High-resolution Surface Electromyography and Computer Vision

Published on: March 28, 2025

Author Spotlight: Addressing Technical and Subjective Challenges in Measuring Classroom Attention

Author Spotlight: Addressing Technical and Subjective Challenges in Measuring Classroom Attention

Published on: December 15, 2023

Area of Science:

Computer Vision
Affective Computing
Human-Computer Interaction

Background:

Facial expressions are widely studied for emotion recognition, but non-verbal body cues are also crucial indicators of affective states.
Psychological and proxemic studies emphasize the body's role in conveying emotions during daily activities.

Purpose of the Study:

To develop a novel method for recognizing affective actions and interactions from videos using full-body, handcrafted features.
To leverage multi-view representation learning and psychological insights for improved emotion recognition.

Main Methods:

Extracted 2D skeletal data from RGB videos to derive multi-view skeleton features.
Employed a bag-of-visual-words approach to model features and generate a codebook for representing affective actions/interactions as histograms.
Utilized Euclidean distance for matching video histograms against a database of class histograms during recognition.

Main Results:

Achieved 93.64% accuracy for affective action recognition and 90.83% for affective interaction recognition on a custom dataset with 6 emotions.
Demonstrated competitive performance against deep learning methods on a public dataset with 6 emotions plus a neutral state.

Conclusions:

The proposed method effectively recognizes affective actions and interactions using handcrafted, full-body features, validating psychological and proxemic findings.
This approach offers a viable alternative to deep learning methods for emotion recognition from body language.