Skeleton-Based Activity Recognition for Children with Autism Using Graph Convolutional Networks
Sensors · 22.07.2026 · 26 (14) · 4638
DOI: 10.3390/s26144638 ↗Kurz gefasst
Fachartikel in der Zeitschrift Sensors, online erschienen am 22.07.2026. Verfasst von Betül Ay und 2 weiteren Autor:innen. Beteiligte Einrichtung: Department of Computer Engineering, Faculty of Engineering, Fırat University, 23119 Elazığ, Türkiye. Hauptthemen laut MEDLINE: Autistic Disorder. Die Arbeit nennt 39 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Themen: Kinderkrankenpflege
Abstract (Originalsprache)
Movement-based and physical activity programs are central tools in autism intervention, so recognizing the activities a child performs during therapy is valuable for objective progress tracking. Manual monitoring of these sessions is time-consuming and subjective, and raw videos raise privacy concerns because it shows identifiable children. We address autism therapeutic activity recognition from privacy-preserving 2D skeletons, and we focus on the practical difficulty of how several therapeutic activities differ only in subtle motion details. As a backbone, we adopt ProtoGCN, a graph convolutional network that represents each action as a combination of learnable motion prototypes. However, this contrastive backbone organizes all classes at once, so it does not enforce a margin between the few pairs that remain entangled after training. We therefore introduce a Refine–Confusable (RC) module, a training-only regularizer that pushes apart the empirically most-confused class pairs using a hinge-margin loss over momentum-updated class centroids. The module changes neither the backbone nor the inference cost. On the MMASD dataset, restricted to the ten-class 2D-skeleton configuration, the RC module improves the base model across random, session-independent, and subject-independent evaluation. The gain is largest on the strictest subject-independent split and a clip-level analysis confirms that this improvement is statistically significant. Under the protocol-matched holdout, the method reaches 96.30% accuracy with 0.959 macro-F1, surpassing recent 2D-skeleton baselines while keeping a lightweight and privacy-preserving modality. The improvements are modest, as expected on a small clinical dataset, and t-SNE and prototype visualizations show that the learned representation is discriminative and interpretable.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Stichwörter der Autor:innen
Autor:innen und Institutionen
- Department of Computer Engineering, Faculty of Engineering, Fırat University, 23119 Elazığ, Türkiye.
- Department of Physical Education and Sports, Middle East Technical University, 06800 Ankara, Türkiye.
ORCID-Profile: Betül Ay
Zitieren
APA
Ay, B., Öztürk, M. A., & Aydın, G. (2026). Skeleton-Based Activity Recognition for Children with Autism Using Graph Convolutional Networks. Sensors, 26(14), 4638. https://doi.org/10.3390/s26144638
Vancouver
Ay B, Öztürk MA, Aydın G. Skeleton-Based Activity Recognition for Children with Autism Using Graph Convolutional Networks. Sensors. 2026;26(14):4638. doi:10.3390/s26144638
BibTeX
@article{Ay2026,
author = {Ay, Betül and Öztürk, Mehmet Ata and Aydın, Galip},
title = {Skeleton-Based Activity Recognition for Children with Autism Using Graph Convolutional Networks},
journal = {Sensors},
year = {2026},
volume = {26},
number = {14},
pages = {4638},
doi = {10.3390/s26144638}
}
EUVIMED