Related Experiment Video
Updated: Jan 6, 2026
06:37
Author Spotlight: Addressing Technical and Subjective Challenges in Measuring Classroom Attention
Published on: December 15, 2023
5.2K
[Emotion detection by motion analysis: a comparison of human and machine-based processing].
Barbara Naor1,2, Dóra Egri1, Angéla Somogyi2
11 Egészségfejlesztési és Sporttudományi Intézet, Trauma és Fájdalom Pszichofiziológiája Kutatócsoport Budapest Magyarország.
Orvosi Hetilap
|November 16, 2025
Summary
Motion analysis shows promise for recognizing emotions through body movements, but no single method is universally best. Combining human observation, classical analysis, and AI offers the most reliable approach for emotion recognition.
Area of Science:
- Psychology and Cognitive Sciences
- Human-Computer Interaction
- Biomedical Engineering
Background:
- Human emotions manifest through various non-verbal cues, including posture, gestures, and movement patterns.
- Motion analysis offers a quantitative method for emotion recognition, complementing traditional approaches.
- Existing research lacks a consensus on the most effective motion analysis techniques for emotion identification.
Purpose of the Study:
- To systematically review motion analysis methods for emotion recognition.
- To compare the efficacy of human-based versus artificial intelligence-based approaches.
- To identify current limitations and future directions in motion-based emotion recognition.
Main Methods:
- Systematic literature review adhering to the PRISMA protocol.
- Comprehensive search across multiple scientific databases yielding 7699 articles.
- Inclusion/exclusion criteria applied, resulting in 16 selected studies for detailed analysis.
Main Results:
- Nine distinct motion analysis methods were identified.
- Four of these methods incorporated artificial intelligence (AI) for emotion recognition.
- Significant variations exist in data processing, technology application, and accuracy among identified methods.
Conclusions:
- No single motion analysis technique is universally accepted for emotion identification.
- AI is a valuable tool but has limitations when used in isolation for emotion recognition.
- Multidisciplinary approaches combining human observation, classical motion analysis, and machine learning show the most reliable results.
