Grand Prize Project of Fairfax County Regional Science Fair, Fourth Place Winner of Regeneron International Science and Engineering Fair.

A-EYE aims to detect diseases in the retina by utilizing multi-stage neural networks. It utilizes a modified-YOLOv3 object detection model, the novel M-Net++ image segmentation model, a basic deep residual network.

A-EYE is a novel approach to retinal disease detection. It uses a multi-stage neural network to detect diseases in the retina. The first stage of the network is a modified-YOLOv3 object detection model that identifies the optic disc and fovea. The second stage is a novel M-Net++ image segmentation model that segments the retinal layers. The final stage is a basic deep residual network that classifies the segmented layers to detect diseases.