The optimization of white blood cell classification based on EfficientNet model

With the quick development of artificial intelligence (AI), it’s common to see that AI has been used in all kinds of industries to improve the efficiency and help us human to pay more attention to other essential aspects. One of the most important applications of AI is to recognize and classify the objects in the picture. Therefore, it can also be utilized in the area of blood test. In this paper, the blood cell recognition system is constructed to identify different kinds of white blood cells. The main structure of this system is a CNN based on EfficientNet, with the optimizing of adjusting parameter like learning rate according to the change of loss and accuracy.