Računalno očitavanje rukom pisanih JMBAG-ova

master's thesis

• Title (english): Computer-based recognition of handwritten student identifier number
• Author: Jakov Ćurić
• Mentor: Marko Čupić (mentor)
• Committee member: Marko Čupić (predsjednik povjerenstva)
• Committee member: Marin Golub (član povjerenstva)
• Committee member: Jan Šnajder (član povjerenstva)
• Granter: University of Zagreb; Faculty of Electrical Engineering and Computing; Zagreb
• Defense date and country: 2024-02-29, Croatia

Abstract

Uvid u različite metode i tehnike za prepoznavanje rukom pisanih znamenki, s naglaskom na segmentaciju i klasifikaciju. Detaljan pregled postojećih tehnika segmentacije i klasifikacije. Posebna pažnja posvećena je pripremi podataka, proširenju trening skupa te problemima s kojima se susrećemo prilikom obrade. Detaljan opis izbora modela, arhitekture te unaprijednog i unatražnjeg prolaza u kontekstu treniranja modela, uključujući inicijalizaciju, optimizaciju, paralelizaciju i pretragu hiperparametara.

Abstract (english)

Insight into various methods and techniques for handwritten digit recognition, with an emphasis on segmentation and classification. A detailed review of existing segmentation and classification techniques. Special attention was given to data preparation, the problems encountered during processing, and the expansion of the training set. A detailed description of the model selection, architecture, and forward and backward passes in the context of model training, including initialization, optimization, parallelization, and hyperparameter search.

• Keywords: duboko učenje; računalni vid; konvolucijske neuronske mreže
• Keywords (english): deep learning; computer vision; convolutional neural networks
• Language: croatian
• URN:NBN: urn:nbn:hr:168:170186
• 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: 2024-10-12 20:22:53