izpis_h1_title_alt

Razvoj aplikacije za virtualno pomerjanje ličil in preurejanje obrazov
ID ZORNADA, LUKA (Author), ID Štruc, Vitomir (Mentor) More about this mentor... This link opens in a new window

.pdfPDF - Presentation file, Download (1,23 MB)
MD5: A67BDBD4A28F7028C186ED2B3CB4D71E

Abstract
V današnjem času se težko izognemo področju digitalizacije. Z napredkom razvoja računalniških sistemov se vedno več področij odloča uporabiti nove možnosti, ki jih predstavlja digitalizacija za izboljšanje obstoječih rešitev ter olajšanje življenja ljudi. V to področje sodi tudi lepotna industrija, ki veliko vlaga v aplikacije za pomerjanje frizur ter ličil. Cilj naloge je ustvariti aplikacijo za doseg prej omenjenih funkcionalnosti, to je urejanje videza slik obrazov, kar vključuje spremembo barve raznih delov glave (las, oči, ustnic, ...) ter dodajanje učinkov ličil. Te funkcionalnosti izvedemo z globokim učenjem za prepoznavo delov glave ter izvedbo ustreznih transformacij nad sliko glede na želene učinke. Tekom dela uporabljamo več modelov za prenos ličil in nato primerjamo kakovost dobljenih rezultatov vsakega modela z anketo. Rezultat dela je funkcionalna aplikacija, ki omogoča uporabniku nalaganje slike obraza in izbiro bodisi spremembe barve poljubnega dela obraza ali priložitvi slike z učinkom ličil, ki ga želi dodati na prvo sliko. Po koncu obdelave se uporabniku prikaže končna slika. Dobljena aplikacija je popolnoma delujoča in vrača zadovoljive rezultate za vsakdanjo rabo.

Language:Slovenian
Keywords:urejanje slik, prenos ličil, razčlenjevanje obraza, globoko učenje
Work type:Undergraduate thesis
Organization:FE - Faculty of Electrical Engineering
Year:2022
PID:20.500.12556/RUL-139569 This link opens in a new window
COBISS.SI-ID:120181763 This link opens in a new window
Publication date in RUL:05.09.2022
Views:769
Downloads:57
Metadata:XML DC-XML DC-RDF
:
Copy citation
Share:Bookmark and Share

Secondary language

Language:English
Title:Development of application for virtual makeup transfer and face editing
Abstract:
Nowadays it is hard to avoid the topic of digitalization. With the advancements of computer systems more and more fields are deciding to encorporate the newly available options presented by digitalization to improve existing solutions and consequently make peoples lives easier. The beauty industry is among these fields, which is heavily investing in applications to try out hairstyles and makeup. The goal of this work is editing the style of facial images, which includes the color of various parts of the head (hair, eyes, lips, ...) and adding makeup effects. We achieve these functionalities with the use of deep learning to differentiate parts of the head and applies a suitable transformation to achieve the desired effect. During the development of the application we use multiple makeup transfer models, the quality of which is later compared in a survey. The result is a functional application, which allows the user to upload an image of a face and select to either change the color of a desired facial part or provide an image with a desired makeup effect, which is then applied to the first image. The result image is then displayed to the user after the process is finished. The resulting application is fully functional and delivers satisfactory results for everyday use.

Keywords:image editing, makeup transfer, face parsing, deep learning

Similar documents

Similar works from RUL:
Similar works from other Slovenian collections:

Back