prikaz prve stranice dokumenta Prepoznavanje znakova primjenom konvolucijskih neuronskih mreža
Access restricted to students and staff of home institution
undergraduate thesis
Prepoznavanje znakova primjenom konvolucijskih neuronskih mreža
Split: University of Split, Faculty of Electrical Engineering, Mechanical Engineering and Naval Architecture, 2021. urn:nbn:hr:179:299299
Gašić, Ognjen
University of Split
Faculty of Electrical Engineering, Mechanical Engineering and Naval Architecture
Department of Electronics and Computing

Institutional repository: Repository of the Faculty of Electrical Engineering, Mechanical Engineering and Naval Architecture

Cite this document

Gašić, O. (2021). Prepoznavanje znakova primjenom konvolucijskih neuronskih mreža (Undergraduate thesis). Split: University of Split, Faculty of Electrical Engineering, Mechanical Engineering and Naval Architecture. Retrieved from https://urn.nsk.hr/urn:nbn:hr:179:299299

Gašić, Ognjen. "Prepoznavanje znakova primjenom konvolucijskih neuronskih mreža." Undergraduate thesis, University of Split, Faculty of Electrical Engineering, Mechanical Engineering and Naval Architecture, 2021. https://urn.nsk.hr/urn:nbn:hr:179:299299

Gašić, Ognjen. "Prepoznavanje znakova primjenom konvolucijskih neuronskih mreža." Undergraduate thesis, University of Split, Faculty of Electrical Engineering, Mechanical Engineering and Naval Architecture, 2021. https://urn.nsk.hr/urn:nbn:hr:179:299299

Gašić, O. (2021). 'Prepoznavanje znakova primjenom konvolucijskih neuronskih mreža', Undergraduate thesis, University of Split, Faculty of Electrical Engineering, Mechanical Engineering and Naval Architecture, accessed 02 December 2024, https://urn.nsk.hr/urn:nbn:hr:179:299299

Gašić O. Prepoznavanje znakova primjenom konvolucijskih neuronskih mreža [Undergraduate thesis]. Split: University of Split, Faculty of Electrical Engineering, Mechanical Engineering and Naval Architecture; 2021 [cited 2024 December 02] Available at: https://urn.nsk.hr/urn:nbn:hr:179:299299

O. Gašić, "Prepoznavanje znakova primjenom konvolucijskih neuronskih mreža", Undergraduate thesis, University of Split, Faculty of Electrical Engineering, Mechanical Engineering and Naval Architecture, Split, 2021. Available at: https://urn.nsk.hr/urn:nbn:hr:179:299299

Metadata
TitlePrepoznavanje znakova primjenom konvolucijskih neuronskih mreža
Title (english)Character recognition using convolutional neural networks
AuthorOgnjen Gašić
MentorMatko Šarić (mentor)
GranterUniversity of Split
Faculty of Electrical Engineering, Mechanical Engineering and Naval Architecture
(Department of Electronics and Computing)
Split
Defense date and country2021-09-21, Croatia
Scientific / art field,
discipline and subdiscipline		TECHNICAL SCIENCES
Computing
Abstract
Ovaj rad započinjemo opisivanjem osnovnih principa strojnog učenja, umjetnih neuronskih mreža te konvolucijskih neuronskih mreža. Nakon toga implementiramo vlastitu programsku izvedbu konvolucijskih neuronskih mreža za prepoznavanje znakova iz MNIST baze podataka korištenjem Keras biblioteke. Gradimo i treniramo osnovni model te za njega stvaramo grafove gubitka i točnosti kroz epohe treniranja. Taj osnovni model zatim evaluiramo. Mijenjanjem broja slojeva i broja filtera po sloju stvaramo nove modele. Nove modele uspoređujemo stvaranjem njihovih grafova točnosti kroz epohe treniranja. Najbolje modele evaluiramo.
Abstract (english)
This thesis starts with description of the basic principles of machine learning, artificial neural networks and convolutional neural networks. After that we implement our own convolutional neural network for recognizing characters from the MNIST database by using the Keras library. We build and train a basic model and create loss and accuracy graphs through training epochs for it. We then evaluate that basic model. By changing the number of layers and the number of filters per layer we create new models. We compare the new models by creating their accuracy graphs through training epochs. We evaluate the best models.
Keywords
Keywords (english)
Languagecroatian
URN:NBNurn:nbn:hr:179:299299
Study programmeTitle: Computing
Study programme type: university
Study level: undergraduate
Academic / professional title: sveučilišni/a prvostupnik/ prvostupnica (baccalaureus/baccalaurea) inženjer/inženjerka računarstva (sveučilišni/a prvostupnik/ prvostupnica (baccalaureus/baccalaurea) inženjer/inženjerka računarstva)
Type of resourceText
File originBorn digital
Access conditionsAccess restricted to students and staff of home institution
Terms of useLicence In copyright icon
Created on2021-12-03 14:15:16