Prepoznavanje rukom pisanih Booleovih izraza na temelju gesti

master's thesis

• Title (english): Gesture Based Recognition of Handwritten Boolean Expressions
• Author: Mate Gašparini
• Mentor: Marko Čupić (mentor)
• Committee member: Marko Čupić (predsjednik povjerenstva)
• Committee member: Alan Jović (član povjerenstva)
• Committee member: Ante Đerek (član povjerenstva)
• Granter: University of Zagreb; Faculty of Electrical Engineering and Computing; Zagreb
• Defense date and country: 2021-07-01, Croatia
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Computing

Abstract

Rukom pisani Booleovi izrazi sastoje se od niza gesti. Prepoznavanje izraza iz niza gesti potrebno je automatizirati radi lakšeg unošenja složenih izraza u memoriju računala. Rješenje izneseno u ovom radu temelji se na samoorganizirajućim i potpuno povezanim unaprijednim neuronskim mrežama učenim genetskim algoritmom. Dodatno, implementirano je mrežno sučelje za prikupljanje podataka za učenje te program za demonstraciju naučenog modela i dodatne vizualizacije. Postupak učenja pokazao se učinkovitim i robusnim u stvaranju modela koji je u većini ispitanih slučajeva dovoljno točan.

Abstract (english)

Handwritten Boolean expressions consist of gesture sequences. Recognition of an expression from a sequence of gestures needs to be automated for the sake of easier input of complex expressions into computer memory. The solution presented in this work is based on self-organizing and fully connected feedforward neural networks using genetic algorithm based training. Furthermore, a web interface for dataset collection has been implemented, as well as a program for learned model demonstration and additional visualizations. The learning process has been proved effective and robust in producing a model that is sufficiently accurate in a majority of tested cases.

• Keywords: neuronske mreže; genetski algoritam; neuroevolucija
• Keywords (english): neural networks; genetic algorithm; neuroevolution
• Language: croatian
• URN:NBN: urn:nbn:hr:168:684928
• Study programme: Title: Computing; Study programme type: university; Study level: graduate; Academic / professional title: magistar/magistra inženjer/inženjerka računarstva (magistar/magistra inženjer/inženjerka računarstva)
• Type of resource: Text
• File origin: Born digital
• Access conditions: Closed access
• Created on: 2023-01-13 12:17:51