2000 character limit reached
Pathological Myopic Image Analysis with Transfer Learning
Published 31 Jul 2019 in eess.IV | (1908.00410v1)
Abstract: We present a summary of transfer learning based methods for several challenging myopic fundus image analysis tasks including classification of pathological and non-pathological myopia,localisation of fovea,and segmentation of optic disc.By adapting existing popular deep learning architectures,our proposed methods have achieved 1st and 2nd place in several tasks at the Pathologic Myopia Challenge held at ISBI2019.
Paper Prompts
Sign up for free to create and run prompts on this paper using GPT-5.
Top Community Prompts
Collections
Sign up for free to add this paper to one or more collections.