Primerjava metod zmanjševanja dimenzionalnosti za napovedovanje deformacij kože 3D modelov

Sekirnik, Rok

| | SLO | ENG | Cookies and privacy

Bigger font | Smaller font

First page > Show document

Show document

Title:	Primerjava metod zmanjševanja dimenzionalnosti za napovedovanje deformacij kože 3D modelov
Authors:	ID Sekirnik, Rok (Author) ID Jesenko, David (Mentor) More about this mentor... ID Bizjak, Marko (Comentor)
Files:	MAG_Sekirnik_Rok_2025.pdf (2,49 MB) MD5: 8618E86AAA1003EC064E2517E3C8F1EF
Language:	Slovenian
Work type:	Master's thesis/paper
Typology:	2.09 - Master's Thesis
Organization:	FERI - Faculty of Electrical Engineering and Computer Science
Abstract:	V magistrski nalogi predstavimo nevronsko mrežo za napoved deformacij kože tridimenzionalnih karakterjev. Osredotočamo se na tehnike redukcije dimenzionalnosti, kot sta metoda glavnih komponent in samokodirnik. Primerjamo zmogljivost obeh pristopov glede natančnosti, uporabnosti in časovne zahtevnosti ter podamo smernice za njihovo praktično uporabo pri generiranju realistično animiranih karakterjev v filmih, računalniški grafiki in realnočasovnih okoljih. Rezultati kažejo, da samokodirnik zagotavlja višjo kakovost napovedi in boljše posploševanje kompleksnih deformacij, medtem ko PCA izstopa po hitrosti in preprosti implementaciji. Ugotovitve potrjujejo, da je izbira metode odvisna od specifičnih zahtev problema, kar omogoča prilagodljiv in učinkovit pristop za različne scenarije v računalniški grafiki.
Keywords:	PCA, samokodirnik, nevronska mreža, fizikalna simulacija, deformacija mreže
Place of publishing:	Maribor
Publisher:	[R. Sekirnik]
Year of publishing:	2025
PID:	20.500.12556/DKUM-94260
UDC:	004.032.26:004.92(043.2)
COBISS.SI-ID:	254190339
Publication date in DKUM:	04.09.2025
Views:	247
Downloads:	28
Metadata:
Categories:	KTFMB - FERI
:	Copy citation

Average score:	(0 votes)
Your score:	Voting is allowed only for logged in users.
Share:

Hover the mouse pointer over a document title to show the abstract or click on the title to get all document metadata.

Licences

License:	CC BY-NC-ND 4.0, Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International

Link:	http://creativecommons.org/licenses/by-nc-nd/4.0/
Description:	The most restrictive Creative Commons license. This only allows people to download and share the work for no commercial gain and for no other purposes.
Licensing start date:	11.08.2025

Secondary language

Language:	English
Title:	Comparing dimensionality reduction methods for predicting skin deformations in 3D models
Abstract:	In this master’s thesis, we present a neural network for predicting skin deformations of three-dimensional characters. We focus on dimensionality reduction techniques, namely Principal Component Analysis and autoencoders. We compare both approaches in terms of accuracy, usability, and computational cost, and provide guidelines for their practical application in generating realistic animated characters for film, computer graphics, and real-time environments. Results indicate that autoencoders offer higher prediction quality and better generalization of complex deformations, whereas PCA excels in speed and simplicity of implementation. Findings confirm that the choice of method depends on specific problem requirements, enabling a flexible and efficient approach for various scenarios in computer graphics.
Keywords:	PCA, autoencoder, neural network, physical simulation, mesh deformation

Comments

Leave comment

You must log in to leave a comment.

Comments (0)

0 - 0 / 0

There are no comments!

Back