Easy-to-use deep learning based super-resolution in microscopy images

Abstract

[ES]El TFG trata de la investigación de una red neuronal reciente de super-resolución así como su implementación en un notebook fácil de usar dirigido a gente no experta en programación.

[EN]The main objective of this work is to present one of the most recent deep neural networks to solve the super-resolution task, named DFCAN, as well as the study of different methods that solve the same problem. Throughout the document, different approaches are explained and compared, emphasising the state-of-the-art methods. This work also contains different experiments done with the DFCAN using different datasets. Finally, to complete the thesis, there is a guide for an easy-to-use Jupyter notebook, created with the aim of being available for anyone, specifically designed for people who do not have expertise in programming and deep learning.