KimCHI: A Sketch-Based Developmental Skill Classifier to Enhance Pen-Driven Educational Interfaces for Children

Bibliography Information:

Hong-hoe Kim, Paul Taele, Stephanie Valentine, Erin McTigue, and Tracy Hammond. 2013. KimCHI: a sketch-based developmental skill classifier to enhance pen-driven educational interfaces for children. In Proceedings of the International Symposium on Sketch-Based Interfaces and Modeling (SBIM '13), Stephen N. Spencer (Ed.). ACM, New York, NY, USA, 33-42. DOI=10.1145/2487381.2487389 http://doi.acm.org/10.1145/2487381.2487389

URL:

http://dl.acm.org/citation.cfm?id=2487389

KimCHI is a sketch recognition system developed toward the recognition of sketches drawn by children. Specifically, it aims to recognize children's sketches and differentiate them between whether the drawer is a pre-school child or an elementary-school level child. Additionally, the way that sketches are drawn can be analyzed to tell apart whether the user is a male or female, and whether the user is an adult or child. The novelty in this recognition system is due to the fact that there has been very little work done on analyzing sketch recognition for children, since children draw shapes very differently depending on their fine motor skills as well as many other factors.

Out of the data from the 725 sketches created by 20 children, classification techniques were used to develop the optimal features for classifying between pre-schoolers vs. grade schoolers. These features are: average curvature of the stroke, direction change ratio, error of the best fit line of the direction strength, and the max curvature to average curvature value. All of these features have a 100% accuracy in determining these factors. The optimal features for classifying between pre-schoolers and adults are Average curvature of the stroke (100%), Direction change ratio (100%), The angle of the major axis relative to center (100%), The error of the best fit line of the direction graph (100%), The maximum curvature to average curvature value (100%), and Slope of the direction graph (100%).

Sketch recognition in children is one of the domains that is highly overlooked. I believe that this is imperative if we want to move sketch recognition as a field into the realm of education. We can use this for number recognition to teach children handwriting as well as basic arithmetic, since these are important pillars of early grade school education.