Thesis (Selection of subject)Thesis (Selection of subject)(version: 336)
Assignment details
   Login via CAS
Generování realistických snímků obloh
Thesis title in Czech: Generování realistických snímků obloh
Thesis title in English: Generation of realistic skydome images
Key words: Hluboké učení, generativní kompetitivní síť, hluboká konvoluční síť, obloha, rybí oko
English key words: Deep learning, generative adversarial network, deep convolutional network, skydome, fisheye
Academic year of topic announcement: 2019/2020
Type of assignment: diploma thesis
Thesis language: čeština
Department: Department of Software and Computer Science Education (32-KSVI)
Supervisor: doc. Alexander Wilkie, Dr.
Author: hidden - assigned and confirmed by the Study Dept.
Date of registration: 03.04.2020
Date of assignment: 03.04.2020
Confirmed by Study dept. on: 29.04.2020
Date and time of defence: 08.07.2020 09:00
Date of electronic submission:20.05.2020
Date of submission of printed version:28.05.2020
Date of proceeded defence: 08.07.2020
Reviewers: Mgr. Martin Pilát, Ph.D.
The goal of the thesis is to generate images of the skydome for use in offline rendering, such as architecture or automotive visualizations, using a deep generative model trained on a large dataset of captured skydome images. The model should produce high-quality images in suitable resolution and the user should be able to control the images using intuitive parameters (season of the year, position of the sun, meteorological conditions). The ultimate aim is to integrate the model into a professional rendering application.
- Karras et al.: Progressive Growing of GANs for Improved Quality, Stability, and Variation (2018),
- Karras et al.: A Style-Based Generator Architecture for Generative Adversarial Networks (2019),
- Karras et al.: Analyzing and Improving the Image Quality of StyleGAN (2019),
- Wang et al.: Deep Learning for Image Super-resolution: A Survey (2019),
- Hosek, Wilkie: An Analytic Model for Full Spectral Sky-Dome Radiance (2012),
- Goodfellow et al.: Deep Learning (2016),
Charles University | Information system of Charles University |