Avatar customization using deep learning’s style transfer technology
Metadata
Show full item recordcomunitat-uji-handle:10234/158176
comunitat-uji-handle2:10234/71324
comunitat-uji-handle3:10234/169451
comunitat-uji-handle4:
TFG-TFMMetadata
Title
Avatar customization using deep learning’s style transfer technologyAuthor (s)
Tutor/Supervisor; University.Department
Fernández Beltrán, Rafael; Universitat Jaume I. Departament de Llenguatges i Sistemes InformàticsDate
2023-07-18Publisher
Universitat Jaume IAbstract
An important aspect in the world of video games is the player's identity, In this regard
player´s avatar plays a significant role in this. However, we often find ourselves limited to a
few predefined options, which ... [+]
An important aspect in the world of video games is the player's identity, In this regard
player´s avatar plays a significant role in this. However, we often find ourselves limited to a
few predefined options, which can restrict our individual expression. This document presents
a project report on an application aiming to address this issue by providing more
customization options to players. This application, developed in tkinter, utilizes a Style
Transfer Model [1] using deep learning techniques (Specifically Convolutional Neural
Networks) to transform user’s self images into a specific artistic style.
Also, it offers a user-friendly interface where users can upload their avatar images or take a
photo, and choose from a predefined list of artistic styles. Once the desired style is selected,
the application applies the style transfer model to generate the transformed image. [-]
Subject
Description
Treball final de Grau en Disseny i Desenvolupament de Videojocs. Codi: VJ1241. Curs acadèmic: 2022/2023
Type
info:eu-repo/semantics/bachelorThesisRights
info:eu-repo/semantics/openAccess