Zpět na vyhledávání

Bc. Michal Čaniga

Bachelor's thesis

Smart Picture Enlargement Using Neural Networks

Smart Picture Enlargement Using Neural Networks
Abstract:
Táto práca je zameraná na inteligentné zvačšovanie fotografií s použitím pokročilých metód syntézy obrazu. Použili sme "super-resolution" GAN sieť implementovanú v Pythone ako fork z knižnice Keras-GAN. Vytvorili sme vlastný malý dataset, na ktorom trénujeme a vyhodnocujeme sieť a porovnávame jej výkon s tradičnými technológiami na zvačšovanie fotografií. Ako evalučné metriky sme použili FID a MSE …more
Abstract:
This thesis focuses on smart picture enlargement using advanced methods of image synthesis. We utilized a super-resolution GAN network implemented in Python as a fork from the Keras-GAN library. We created a small custom dataset on which we train and evaluate the network and compare its performance with traditional techniques for image enlargement. As evaluation metrics, we pick MSE and FID.
 

Keywords

CBIA deep learning GPU image processing machine learning neural networks image upscaling
 
Language used: English
Date on which the thesis was submitted / produced: 25. 5. 2021
Identifier: https://is.muni.cz/th/counf/

Thesis defence

  • Date of defence: 30. 6. 2021
  • Supervisor: RNDr. David Wiesner
  • Reader: doc. RNDr. Martin Maška, Ph.D.

Citation record

Cite this text

Full text of thesis

Contents of on-line thesis archive
Published in Theses:
  • světu
Other ways of accessing the text
Institution archiving the thesis and making it accessible: Masarykova univerzita, Fakulta informatiky
Složka Reference to the local database directory of the institution

Masaryk University

Faculty of Informatics

Bachelor programme / field:
Applied Informatics / Applied Informatics

Theses on a related topic

List of theses with an identical keyword.