Retinal Blood Vessel Segmentation Using a Fully Convolutional Network - Transfer Learning from Patch- to Image-Level.
Taibou Birgui SekouMoncef HidaneJulien OlivierHubert CardotPublished in: MLMI@MICCAI (2018)
Keyphrases
- blood vessels
- transfer learning
- retinal fundus images
- image analysis
- pixel level
- segmentation method
- image segmentation
- segmentation algorithm
- multiscale
- vessel detection
- test images
- input image
- convolutional network
- region of interest
- edge detection
- retinal images
- image classification
- collaborative filtering
- high resolution
- optic disc
- reinforcement learning
- learning algorithm
- convolutional neural networks
- energy function
- image retrieval
- image content
- labeled data
- text classification
- level set
- active learning
- image collections
- machine learning
- cross domain
- recognition scheme
- text mining
- image registration
- computer vision
- information retrieval