What are the main retina datasets?
A select number of retinal datasets are proposed.
While retina images are harder to capture than face or fingerprints, there’s a selection of them as retinal photography is used in medicine for diagnosing ocular (and some other) diseases.
Messidor and Messidor-2 datasets are two well-known examples assembled at the research center Mines Paris for diagnosing diabetic retinopathy. They provided paired fundus images supplied with third-party annotations.
Retina Identification Database (RIDB) was created with a TOPCON-TRC camera and offers 100 samples from 20 people. PAPILA is a public funds dataset with segmentations of the cup and optic disc. All can be used for retinal anti-spoofing as well.