Semantic Segmentation of Diabetic Retinopathy Lesions, Using a UNET with Pretrained Encoder.
Dimitrios TheodoropoulosGeorgios C. ManikisKostas MariasGiorgos PapadourakisPublished in: EANN (2022)
Keyphrases
- diabetic retinopathy
- semantic segmentation
- automatic detection
- retinal images
- early detection
- superpixels
- label transfer
- conditional random fields
- bit rate
- medical experts
- weakly supervised
- scene classification
- blood vessels
- object categories
- optic disc
- object class
- object classes
- vessel segmentation
- pascal voc
- machine learning
- image understanding
- image set
- higher order
- bag of words
- semi supervised
- probabilistic model
- multiscale
- image segmentation