ECA-CBAM: Classification of Diabetic Retinopathy: Classification of diabetic retinopathy by cross-combined attention mechanism

Abstract

Although there is no distinctive header, this is the abstract. Diabetic retinopathy is an ophthalmological disease that causes bleeding in the fundus and loss of vision due to damage to blood vessels in the retina. It is one of the main causes of vision loss in the world. To slow down the development of the disease, early screening of the eyeball is needed. This paper proposes a new method of classification, automatic screening and accurate diagnosis of diabetic retinopathy based on convolutional neural network. Specifically, five attention mechanisms such as BAM, CBAM, ECA, CA and SeNet are used to classify diabetic retinopathy. Through comparative experiments, it is found that ECA-CBAM cross-combination model has the best classification performance.