Age detection using clustering algorithms from children's drawings DOI Creative Commons

Messaoud Mezati,

Ines Aouria

STUDIES IN ENGINEERING AND EXACT SCIENCES, Journal Year: 2024, Volume and Issue: 5(2), P. e10582 - e10582

Published: Nov. 13, 2024

Age detection from children's drawings is an innovative approach to understanding developmental milestones through visual analysis. Traditional methods for determining a child's age often rely on linguistic or cognitive assessments, but data such as offer untapped potential non-invasive This study explores the use of clustering algorithms detect patterns in drawings, providing novel method estimation. A dataset representing different groups, was collected, and key features, line thickness, object proportions, were extracted. These features analyzed using unsupervised algorithms, including K-means, Agglomerative Clustering, Mean-shift, others, group based age-related characteristics. Among tested, Mean-shift achieved highest performance, with silhouette score 0.67 mapping clusters correct labels. K-means Clustering exhibited moderate scores 0.57 0.46. In contrast, Spectral OPTICS performed poorly, negative scores, reflecting poorly defined cluster boundaries. Our demonstrates automatic despite challenges overlapping between adjacent groups. The findings suggest directions future research, more complex models that integrate indicators enhanced assessment. has significant implications educational psychology, child development, artificial intelligence.

Language: Английский

Age detection using clustering algorithms from children's drawings DOI Creative Commons

Messaoud Mezati,

Ines Aouria

STUDIES IN ENGINEERING AND EXACT SCIENCES, Journal Year: 2024, Volume and Issue: 5(2), P. e10582 - e10582

Published: Nov. 13, 2024

Age detection from children's drawings is an innovative approach to understanding developmental milestones through visual analysis. Traditional methods for determining a child's age often rely on linguistic or cognitive assessments, but data such as offer untapped potential non-invasive This study explores the use of clustering algorithms detect patterns in drawings, providing novel method estimation. A dataset representing different groups, was collected, and key features, line thickness, object proportions, were extracted. These features analyzed using unsupervised algorithms, including K-means, Agglomerative Clustering, Mean-shift, others, group based age-related characteristics. Among tested, Mean-shift achieved highest performance, with silhouette score 0.67 mapping clusters correct labels. K-means Clustering exhibited moderate scores 0.57 0.46. In contrast, Spectral OPTICS performed poorly, negative scores, reflecting poorly defined cluster boundaries. Our demonstrates automatic despite challenges overlapping between adjacent groups. The findings suggest directions future research, more complex models that integrate indicators enhanced assessment. has significant implications educational psychology, child development, artificial intelligence.

Language: Английский

Citations

0