: It provides three levels of spatial annotations: binary vessel masks, general vein-artery masks, and fine-grained vein-artery masks.
: 635 smartphone-based videos from four different clinics, featuring patients aged 50–75.
The search result for typically refers to the Handheld Device-Based Fundus Video Dataset (RVD) , which is the official repository for the paper titled " Leveraging Frame Affinity for Retinal Vessel Segmentation in Fundus Videos " published at CVPR 2024.