Análisis comparativo de clasificadores para gestos descritos por LaGeR

Abstract

The present thesis is a comparative analysis between the Damerau-Levenshtein algorithm and a deep learning algorithm and con rms that, by using deep learning techniques, one can obtain better results when recognizing gestures described by LaGeR, in terms of both running time and accuracy. Additionally, it shows that said method is e ective with several input devices. xvii