This paper describes a trainable recognizer for hand-drawn sketches using geometric features. We compare three different learning algorithms and select the best approach in terms of cost-performance ratio. The algorithms employ classic machine-learning techniques using a clustering approach. Experimental results show competing performance (95.1%) with the non-trainable recognizer (95.8%) previously developed, with obvious gains in flexibility and expandability. In addition, we study both their classification and learning performance with increasing number of examples per class. © Springer-Verlag Berlin Heidelberg 2002.
CITATION STYLE
Pimentel, C. F., da Fonseca, M. J., & Jorge, J. A. (2002). Experimental evaluation of a trainable scribble recognizer for calligraphic interfaces. Lecture Notes in Computer Science (Including Subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 2390, 81–91. https://doi.org/10.1007/3-540-45868-9_7
Mendeley helps you to discover research relevant for your work.