Napovedovanje obrazov otrok z GAN metodo

diplomsko delo

Nejc Šuklje (Author), Luka Šajn (Mentor)

Abstract

Diplomska naloga z naslovom ”Napovedovanje slik otrok obrazov z GAN metodo”ˇze na prvi pogled vzbudi motivacijo, ker je potencialni rezultat zanimiv. GAN metoda je tudi dokaj nova stvar, kar vse skupaj naredi ˇse bolj zanimivo, saj ni veliko raziskav, v katerih je uporabljena. Na problem lahko pogledamo tudi s strani ”kupca”, torej nekoga, ki bi rad videl sliko svojega potencialnega otroka. Po mojem mnenju stvar, ki bi na trgu doˇzivela porast, ˇse posebej v bogatejˇsem sloju prebivalstva. Ceprav je raziskav na to temo malo, bodo ˇ vseeno pomagala pri izdelavi naloge. Morda toliko bolj, ker bo moˇzno bolj podrobno predelati podroˇcje. Programerski del bo napisan v programskem jeziku python. Na koncu bomo analizirali rezultate in naredili primerjavo z dejanskimi slikami obrazov otrok in ocenili uspeh. Poiskali bomo tudi toˇcke, ki bi se jih dalo izboljˇsati, da bi bil izdelek ˇse boljˇsi in uporabnejˇsi.

Keywords

grebenska regresija;latentni prostor;GAN;generiranje obrazov;univerzitetni študij;diplomske naloge;

Data

Language:	Slovenian
Year of publishing:	2022
Typology:	2.11 - Undergraduate Thesis
Organization:	UL FRI - Faculty of Computer and Information Science
Publisher:	[N. Šuklje]
UDC:	004.932(043.2)
COBISS:	135339267
Views:	33
Downloads:	10
Average score:	0 (0 votes)
Metadata:

Other data

Secondary language:	English
Secondary title:	Predicting children faces with GAN method
Secondary abstract:	The diploma thesis entitled ”Prediction of images of children’s faces using the GAN method” arouses motivation at first sight, because the potential result is interesting. The GAN method is also a fairly new thing, which makes it all the more interesting because there is not much research in which it is used. We can also look at the problem from the side of the ”buyer”, that is someone who would like to see a picture of their potential child. In my opinion, a thing that would see an increase in the market, especially in the wealthier population. Although there is little research on this topic, they will still help in making the assignment. Perhaps even more so because it will be possible to rework the area in more detail. The programming part will be written in the python programming language. Finally, we will analyze the results and make a comparison from real images of children and evaluate the success. We will also look for points that could be improved to make the product even better and more useful.
Secondary keywords:	ridge regression;latent space;GAN;generation of faces;computer science;diploma;Računalništvo;Univerzitetna in visokošolska dela;
Type (COBISS):	Bachelor thesis/paper
Study programme:	1000468
Embargo end date (OpenAIRE):	1970-01-01
Thesis comment:	Univ. v Ljubljani, Fak. za računalništvo in informatiko
Pages:	28 str.
ID:	17491047