Generiranje slika generativnim suparničkim mrežama

master's thesis

• Title (english): Generating Images Using Generative Adversarial Networks
• Author: Bruno Zebić
• Mentor: Tomislav Hrkać (mentor)
• Committee member: Tomislav Hrkać (predsjednik povjerenstva)
• Committee member: Zoran Kalafatić (član povjerenstva)
• Committee member: Siniša Šegvić (član povjerenstva)
• Granter: University of Zagreb; Faculty of Electrical Engineering and Computing; Zagreb
• Defense date and country: 2021-07-16, Croatia
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Computing

Abstract

U okviru ovog rada proučeni su najznačajniji generativni suparnički učeni modeli opisani u literature i programski je ostvaren takav sustav temeljen na prikladnom modelu i uz korištenje prikladnih biblioteka.
Proučena su četiri značajna generativna suparnička modela i jedan od njih je odabran kao prikladan model za ostvarivanje zadatka.
Pripremljena je baza slika realističnih fotografija lica ljudi na kojima bi se sustav mogao učiti i ispitivati.
Sustav je zatim implementiran.
Njegovi rezultati su analizirani i prikazani grafički. Rezultat je takav da je dio slika uspješno imitirao fotografije realističnih slika ljudi.

Abstract (english)

Within this paper, the most significant generative adversarial learning models described in the literature were studied and such a system based on a suitable model and with the use of appropriate libraries was programmatically realized.
Four significant generative adversarial models were studied and one of them was selected as a suitable model for accomplishing the task.
A database of images of realistic photographs of people's faces has been prepared on which the system could be learned and tested.
The system is then implemented.
Its results were analyzed and presented graphically. The result is such that some of the images successfully mimicked photographs of realistic images of people.

• Keywords: neuronske mreže; slike; lica; generativne suparničke mreže
• Keywords (english): neural networks; images; faces; generative adversarial networks
• Language: croatian
• URN:NBN: urn:nbn:hr:168:312068
• 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: 2023-01-13 14:17:42