prikaz prve stranice dokumenta Artify: prijenos stila slike u stvarnom vremenu s dubokim učenjem
No public access
master's thesis
Artify: prijenos stila slike u stvarnom vremenu s dubokim učenjem
Zagreb: University of Zagreb, Faculty of Electrical Engineering and Computing, 2023. urn:nbn:hr:168:270943
Petešić, Lovre
University of Zagreb
Faculty of Electrical Engineering and Computing

Institutional repository: FER Repository

Cite this document

Petešić, L. (2023). Artify: prijenos stila slike u stvarnom vremenu s dubokim učenjem (Master's thesis). Zagreb: University of Zagreb, Faculty of Electrical Engineering and Computing. Retrieved from https://urn.nsk.hr/urn:nbn:hr:168:270943

Petešić, Lovre. "Artify: prijenos stila slike u stvarnom vremenu s dubokim učenjem." Master's thesis, University of Zagreb, Faculty of Electrical Engineering and Computing, 2023. https://urn.nsk.hr/urn:nbn:hr:168:270943

Petešić, Lovre. "Artify: prijenos stila slike u stvarnom vremenu s dubokim učenjem." Master's thesis, University of Zagreb, Faculty of Electrical Engineering and Computing, 2023. https://urn.nsk.hr/urn:nbn:hr:168:270943

Petešić, L. (2023). 'Artify: prijenos stila slike u stvarnom vremenu s dubokim učenjem', Master's thesis, University of Zagreb, Faculty of Electrical Engineering and Computing, accessed 28 December 2024, https://urn.nsk.hr/urn:nbn:hr:168:270943

Petešić L. Artify: prijenos stila slike u stvarnom vremenu s dubokim učenjem [Master's thesis]. Zagreb: University of Zagreb, Faculty of Electrical Engineering and Computing; 2023 [cited 2024 December 28] Available at: https://urn.nsk.hr/urn:nbn:hr:168:270943

L. Petešić, "Artify: prijenos stila slike u stvarnom vremenu s dubokim učenjem", Master's thesis, University of Zagreb, Faculty of Electrical Engineering and Computing, Zagreb, 2023. Available at: https://urn.nsk.hr/urn:nbn:hr:168:270943

Metadata
TitleArtify: prijenos stila slike u stvarnom vremenu s dubokim učenjem
Title (english)Artify: Real-time Image Style Transfer with Deep Learning
AuthorLovre Petešić
MentorMarina Bagić Babac (mentor)
Committee memberMarina Bagić Babac (predsjednik povjerenstva)
Committee memberMirjana Domazet-Lošo (član povjerenstva)
Committee memberMarko Horvat (član povjerenstva)
GranterUniversity of Zagreb
Faculty of Electrical Engineering and Computing
Zagreb
Defense date and country2023-07-10, Croatia
Scientific / art field,
discipline and subdiscipline		TECHNICAL SCIENCES
Computing
Abstract
Implementirana su dva modela dubokog učenja, od kojih jedan nije uspio biti treniran jer je set za treniranje bio loše odabran. Iz dijelova propalog modela se pokazalo da se maska za isključivanje piksela može izvesti jednostavnije nego što je navedeno u proučenoj literaturi, uz cijenu dodatnog hiperparametra.Pokazalo se i da je TVR gubitak jako osjetljiv na hiperparametre te da je ispravan odabir podataka jedan od ključnih dijelova treniranja neuronskih mreža. Nad uspješnim modelom, modificirali su se parametri poput broja kanala, epoha treniranja i slično. Pokazalo se da je TVR manje potreban ako se ne koristi transponirana konvolucija, već interpolacija za uvećavanje rezolucije slike. Interpolacija sprječava pojavu šahovničkih uzoraka u pikselima pri uvećavanju slike.
Abstract (english)
Two deep learning models were implemented, of which one was unable to be properly trained because the training datased was poorly picked. From parts of the failed model, it was shown that the disocclusion mask can be calculated more easily than shown in researched literature, albeit at a cost of an extra hyperparameter- It was also shown that the TVR loss i svery sensitive to hiperparameters and that attention to correct dana collection and processing is one of the most important aspects of training any neural network. Over the successful model modifications were made like changing the channel size, epochs etc. It was shown that the TVR loss is less needed if one does not use transposed convolution, but interpolation upscaling instead. Interpolation prevents the appearance of checkerboard patterns when upscaling.
Keywords
Keywords (english)
Languageenglish
URN:NBNurn:nbn:hr:168:270943
Study programmeTitle: Computing
Study programme type: university
Study level: graduate
Academic / professional title: magistar/magistra inženjer/inženjerka računarstva (magistar/magistra inženjer/inženjerka računarstva)
Type of resourceText
File originBorn digital
Access conditionsClosed access
Terms of useLicence In copyright icon
Public note
Created on2023-10-11 10:07:53