Related Experiment Video
Updated: Oct 2, 2025
05:32
Comparing Eye-tracking Data of Children with High-functioning ASD, Comorbid ADHD, and of a Control Watching Social Videos
Published on: December 7, 2018
9.1K
Automatic Emotion Recognition in Children with Autism: A Systematic Literature Review
Agnieszka Landowska1, Aleksandra Karpus1, Teresa Zawadzka1
1Faculty of Electronics, Telecommunications and Informatics, Gdańsk University of Technology, 80-233 Gdańsk, Poland.
Sensors (Basel, Switzerland)
|February 26, 2022
Summary
This study reviews methods for automatic emotion recognition in children, crucial for autism therapy. It highlights challenges in data collection and analysis for accurate emotion detection.
Area of Science:
- Psychology
- Computer Science
- Artificial Intelligence
Background:
- Automatic emotion recognition (AER) offers potential advancements for autism therapy.
- Children with autism spectrum disorder (ASD) present unique challenges in emotion recognition.
Purpose of the Study:
- To explore methods and tools for recognizing emotions in children, particularly those with autism.
- To systematically review existing literature on AER in children using PRISMA methodology.
Main Methods:
- Systematic literature review adhering to PRISMA guidelines.
- Analysis of diverse observation channels: facial expressions, speech prosody, physiological signals.
- Evaluation of representation models, classification algorithms (SVM, neural networks), and fusion techniques (early fusion).
Main Results:
- Basic emotions (happiness, fear, sadness) are most frequently recognized.
- Single-channel approaches are preferred over multichannel ones.
- SVM and neural networks are popular for classification; early fusion is common in multimodal recognition.
- Challenges identified include stimuli selection, labeling, and dataset creation; all channels are prone to disturbances.
Conclusions:
- AER methods show promise for enhancing autism therapy.
- Further research is needed to address challenges in data acquisition, standardization, and dataset availability for robust emotion recognition in children.
- Developing reliable AER systems requires careful consideration of participant characteristics and multimodal data integration.
