融入多尺度双线性注意力的轻量化眼底疾病多分类网络

现在大多数眼疾分类方法都是对单一类别疾病不同级别进行分类，并且网络模型存在参数量大、计算复杂等问题。为解决这些问题，提出一种轻量化的眼底疾病多分类网络MELCNet，该网络以PPLCNet为主干网络，由输入层特征提取、并行多尺度结构、双线性SE注意力模块、深度可分离卷积、更小参数计算的h-swish激活函数构成，能关注到不同尺度不同疾病的关键患病信息。实验结果表明，提出的多尺度注意力轻量网络模型具有较少的参数量和计算复杂度，并在所选的四种眼底疾病和正常眼底图像的多分类上取得了优异的分类结果，在内部组合数据集测试集上的分类准确率相对于ResNet-50提升1.11%，相对于Xie等人提出的类似眼疾多分类网络在公开数据集cataract测试集上提升2.5%，相较于其他轻量级分类网络在眼底疾病多分类领域具有较高的准确率以及较强的鲁棒性。

Lightweight fundus disease multi-classification network with multi-scale bilinear attention

Most of the eye disease classification methods are to classify a single category of diseases at different levels, and the network model has problems such as large parameters and complex calculations. To solve these problems, this paper proposed a lightweight fundus disease multi-classification network, MELCNet in this paper. The network used PPLCNet as the backbone, which was composed of input layer for feature extraction, parallel multi-scale structure, bilinear SE attention module, depth separable convolution and the h-swish activation function which calculated with smaller parameters, in order to pay attention to the key disease information of different scales and different diseases. Experimental results show that the proposed multi-scale attention lightweight network model has less parameters and computational complexity, and excellent classification results have been obtained in the multi-classification of the four selected fundus diseases and normal fundus images. The classification accuracy in the internal combined test set is 1.11% higher than that in ResNet-50, and the performance on the public datasets is improved by 2.5% relative to the multi-classification literature of similar type network by Xie, and compared with other lightweight classification networks, it has higher accuracy and strong robustness in the field of multi-classification of fundus diseases.