Eijkelenkamp, Tom (2023) Generative AI Art - Finding a metric to evaluate light illumination. Master's Internship Report, Computing Science.
|
Text
Generative_AI_Art_Internship.pdf Download (36MB) | Preview |
|
![[img]](https://fse.studenttheses.ub.rug.nl/style/images/fileicons/text.png) |
Text
toestemming.pdf Restricted to Registered users only Download (133kB) |
Abstract
For the lately very popular text to image synthesis we explore various methods to find a metric to evaluate realism regarding light illumination within a generated image. First we look at the architecture that drives the photo synthesis, specifically focussing on diffusion. As a method we explore if FID score can be modified into a metric evaluating light interaction within a picture. Additionally we study if a 3D representation can be computed of simplistic scene generated photos. Using this as a ground truth and comparing it against the original generated photo gives us an error value, which potentially can serve as another metric to evaluate light illumination.
| Item Type: | Thesis (Master's Internship Report) |
|---|---|
| Supervisor name: | Wilkinson, M.H.F. |
| Degree programme: | Computing Science |
| Thesis type: | Master's Internship Report |
| Language: | English |
| Date Deposited: | 14 Feb 2024 14:40 |
| Last Modified: | 14 Feb 2024 14:40 |
| URI: | https://fse.studenttheses.ub.rug.nl/id/eprint/31954 |
Actions (login required)
 |
View Item |