Gesture recognition for sentence synthesis with probabilistic algorithmsGesture recognition for sentence synthesis with probabilistic algorithmsΑναγνώριση χειρονομιών για σύνθεση προτάσεων με πιθανοτικούς αλγόριθμους Διπλωματική Εργασία Diploma Work 2018-07-132018enThe purpose of this work was to develop a standalone system that predicts text and outputs speech synthesis for people with physical disabilities and reduced fine motor function. A low-cost, infrared hand tracking device was utilized, typically used in virtual/augmented reality applications. The implemented system offers two basic functionalities. First, a human-computer interaction (HCI) functionality that is customizable and offers enough features that can be tailored to a large number of different users. Second, a way to predict what the user intends to write in order to reduce the user’s input text as much as possible, speeding up overall communication. The word prediction task was mainly based on n-grams (and relevant Markov Chains), abbreviation expansion (and lookup tables) and recurrent neural networks. Evaluation results confirmed that the simple prediction methods utilized accelerated word typing speed, saving input text by approximately 63%. Future work could utilize additional eye-tracking or hand-tracking sensors.http://creativecommons.org/licenses/by-nc-sa/4.0/Πολυτεχνείο Κρήτης::Σχολή Ηλεκτρολόγων Μηχανικών και Μηχανικών ΥπολογιστώνGiannelos_Evangelos_Dip_2018.pdfChania [Greece]Library of TUC2018-07-13application/pdf10.5 MBfree Giannelos Evangelos Γιαννελος Ευαγγελος Bletsas Aggelos Μπλετσας Αγγελος Mania Aikaterini Μανια Αικατερινη Partsinevelos Panagiotis Παρτσινεβελος Παναγιωτης Πολυτεχνείο Κρήτης Technical University of Crete Natural Language Processing Human-Computer Interaction Augmented and Alternative Communication