prikaz prve stranice dokumenta Nenadzirano učenje procjene dubine jednookim vidom
No public access
master's thesis
Nenadzirano učenje procjene dubine jednookim vidom
Zagreb: University of Zagreb, Faculty of Electrical Engineering and Computing, 2019. urn:nbn:hr:168:512297
Bartol, Kristijan
University of Zagreb
Faculty of Electrical Engineering and Computing

Institutional repository: FER Repository

Cite this document

Bartol, K. (2019). Nenadzirano učenje procjene dubine jednookim vidom (Master's thesis). Zagreb: University of Zagreb, Faculty of Electrical Engineering and Computing. Retrieved from https://urn.nsk.hr/urn:nbn:hr:168:512297

Bartol, Kristijan. "Nenadzirano učenje procjene dubine jednookim vidom." Master's thesis, University of Zagreb, Faculty of Electrical Engineering and Computing, 2019. https://urn.nsk.hr/urn:nbn:hr:168:512297

Bartol, Kristijan. "Nenadzirano učenje procjene dubine jednookim vidom." Master's thesis, University of Zagreb, Faculty of Electrical Engineering and Computing, 2019. https://urn.nsk.hr/urn:nbn:hr:168:512297

Bartol, K. (2019). 'Nenadzirano učenje procjene dubine jednookim vidom', Master's thesis, University of Zagreb, Faculty of Electrical Engineering and Computing, accessed 26 December 2024, https://urn.nsk.hr/urn:nbn:hr:168:512297

Bartol K. Nenadzirano učenje procjene dubine jednookim vidom [Master's thesis]. Zagreb: University of Zagreb, Faculty of Electrical Engineering and Computing; 2019 [cited 2024 December 26] Available at: https://urn.nsk.hr/urn:nbn:hr:168:512297

K. Bartol, "Nenadzirano učenje procjene dubine jednookim vidom", Master's thesis, University of Zagreb, Faculty of Electrical Engineering and Computing, Zagreb, 2019. Available at: https://urn.nsk.hr/urn:nbn:hr:168:512297

Metadata
TitleNenadzirano učenje procjene dubine jednookim vidom
Title (english)Unsupervised learning of depth estimation with monocular vision
AuthorKristijan Bartol
MentorSiniša Šegvić (mentor)
Committee memberSiniša Šegvić (predsjednik povjerenstva)
Committee memberZoran Kalafatić (član povjerenstva)
Committee memberTomislav Hrkać (član povjerenstva)
GranterUniversity of Zagreb
Faculty of Electrical Engineering and Computing
Zagreb
Defense date and country2019-03-01, Croatia
Scientific / art field,
discipline and subdiscipline		TECHNICAL SCIENCES
Computing
Abstract
Rekonstrukcija strukture prirodnih scena vrlo je važan sastojak mnogih praktičnih primjena računalnog vida. Posebno su zanimljivi pristupi koji učenje provode na slijedu slika jedne kamere jer ne zahtijevaju komplicirane sustave za pribavljanje i označavanje podataka za učenje. U posljednje vrijeme, veliki uspjeh u tom području ostvaruju metodu temeljene na dubokim konvolucijskim modelima. U okviru rada, proučen je model nenadziranog učenja procjene geometrijskih značajki iz monokularnog niza slika. Opisana je implementacija dubokog modela i prikazani su rezultati dubinskih mapa dobivenih treniranim modelom.
Abstract (english)
Natural scene reconstruction is a very important ingredient for a lot of practical application in computer vision. Learning from sequence of monocular images is particularly interesting as they don't demand complicated systems for fetching and labeling the dataset. Recently, a grand success is achieved using methods based on deep convolutional networks. Unsupervised model for learning scene geometry from monocular image sequence is analyzed in this work. The implementaion of deep model is described and the depth maps, generated using the model, are shown.
Keywords
Keywords (english)
Languagecroatian
URN:NBNurn:nbn:hr:168:512297
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 on2019-03-13 18:53:23