Build and train advanced model for image generation
Lin, Meng
This item is only available for download by members of the University of Illinois community. Students, faculty, and staff at the U of I may log in with your NetID and password to view the item. If you are trying to access an Illinois-restricted dissertation or thesis, you can request a copy through your library's Inter-Library Loan office or purchase a copy directly from ProQuest.
Permalink
https://hdl.handle.net/2142/97867
Description
Title
Build and train advanced model for image generation
Author(s)
Lin, Meng
Contributor(s)
Lazebnik, Svetlana
Issue Date
2017-05
Keyword(s)
advanced model of image generation
VAE-GAN model
Variational Autoencoder (VAN) Generative Adversarial Network model
Abstract
Recent developments on deep learning have enabled generative models to capture distribution of
relatively complex datasets. In this research we aim to build a
cutting-edge model that is able to
learn the distribution of the CelebA face dataset. We surveyed several papers published in recent years
and decided to construct and improve the VAE-GAN model, which combines the Variational Autoencoder (VAE)
and Generative Adversarial Network (GAN). Chapter 1 introduces GAN, VAE, VAE-GAN model and
addresses a disadvantage of GAN and proposes some methods to improve it. We also introduce
Adversarial Autoencoder which we intend to accompany the model in the future. The other chapters
address some details of implementation and show the results of the experiment.
Use this login method if you
don't
have an
@illinois.edu
email address.
(Oops, I do have one)
IDEALS migrated to a new platform on June 23, 2022. If you created
your account prior to this date, you will have to reset your password
using the forgot-password link below.
