prikaz prve stranice dokumenta Učenje semantičke segmentacije na nepotpunim oznakama
No public access
master's thesis
Učenje semantičke segmentacije na nepotpunim oznakama
2023. urn:nbn:hr:168:926062
Borko, Petar
University of Zagreb
Faculty of Electrical Engineering and Computing

Cite this document

Borko, P. (2023). Učenje semantičke segmentacije na nepotpunim oznakama (Master's thesis). Zagreb: University of Zagreb, Faculty of Electrical Engineering and Computing. Retrieved from https://urn.nsk.hr/urn:nbn:hr:168:926062

Borko, Petar. "Učenje semantičke segmentacije na nepotpunim oznakama." Master's thesis, University of Zagreb, Faculty of Electrical Engineering and Computing, 2023. https://urn.nsk.hr/urn:nbn:hr:168:926062

Borko, Petar. "Učenje semantičke segmentacije na nepotpunim oznakama." Master's thesis, University of Zagreb, Faculty of Electrical Engineering and Computing, 2023. https://urn.nsk.hr/urn:nbn:hr:168:926062

Borko, P. (2023). 'Učenje semantičke segmentacije na nepotpunim oznakama', Master's thesis, University of Zagreb, Faculty of Electrical Engineering and Computing, accessed 12 July 2024, https://urn.nsk.hr/urn:nbn:hr:168:926062

Borko P. Učenje semantičke segmentacije na nepotpunim oznakama [Master's thesis]. Zagreb: University of Zagreb, Faculty of Electrical Engineering and Computing; 2023 [cited 2024 July 12] Available at: https://urn.nsk.hr/urn:nbn:hr:168:926062

P. Borko, "Učenje semantičke segmentacije na nepotpunim oznakama", Master's thesis, University of Zagreb, Faculty of Electrical Engineering and Computing, Zagreb, 2023. Available at: https://urn.nsk.hr/urn:nbn:hr:168:926062

Please login to the repository to save this object to your list.
Metadata
TitleUčenje semantičke segmentacije na nepotpunim oznakama
Title (english)Training semantic segmentation on incomplete labels
AuthorPetar Borko
MentorSiniša Šegvić (mentor)
Committee memberSiniša Šegvić (predsjednik povjerenstva)
Committee memberMarko Čupić (član povjerenstva)
Committee memberLeonardo Jelenković (član povjerenstva)
GranterUniversity of Zagreb
Faculty of Electrical Engineering and Computing
Zagreb
Defense date and country2023-07-08, Croatia
Abstract
Semantička segmentacija važan je zadatak računalnog vida koji zahtjeva veliku količinu označenih podataka sa kojima često ne raspolaže. Iz tog razloga su smišljene tehnike polunadziranog strojnog učenja koje bi iskoristile moć neoznačenih ili slabo označenih podataka kako bi stvorili induktivne pretpostavke o čitavoj distribuciji. Ovaj rad bavi se uporabom kombinacije postojećih modela kako bi se ostvarili bolji rezultati na osobnom skupu podataka. Demonstriraju se moderni alati i tehnike učenja kako bi se ostvarila zadovoljavajuća točnost uz pomoć snage neoznačenih podataka.
Abstract (english)
Semantic segmentation is an important computer vision task that requires a great deal of annotated data, which is seldom the case. For this reason, weakly supervised techniques have been developed for leveraging the power of unlabeled and weakly labeled data to make inductive reasoning about the underlying data distribution. This work is based on using a combination of powerful models to achieve better results on a custom dataset. Modern pseudolabeling techniques are used to leverage the power of unlabeled and weakly annotated data to achieve sufficient accuracy.
Keywords
Keywords (english)
Languagecroatian
URN:NBNurn:nbn:hr:168:926062
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 08:29:12