References & Citations
Computer Science > Computer Vision and Pattern Recognition
Title: Deep Learning for Cornea Microscopy Blind Deblurring
(Submitted on 25 Jun 2020)
Abstract: The goal of this project is to build a deep-learning solution that deblurs cornea scans, used for medical examination. The spherical shape of the eye prevents ophtamologist from having completely sharp image. Provided with a stack of corneas from confocal images, our approach is to build a model that performs an upscaling of the images using an SR (Super Resolution) Network.
Link back to: arXiv, form interface, contact.