Image identification model training method and device, and image identification method and device

Abstract

The invention discloses an image identification model training method. The method comprises the steps of obtaining a to-be-trained image set; obtaining a first prediction probability, a second prediction probability, a third prediction probability and a fourth prediction probability through a to-be-trained image recognition model based on the to-be-trained image set; determining a target loss function according to the first prediction probability, the second prediction probability, the third prediction probability and the fourth prediction probability; and training the to-be-trained image recognition model based on the target loss function to obtain an image recognition model. The invention further discloses an image recognition method and a device. According to the method, annotation is adopted; the model is trained for the medical images of different tasks and the unlabeled medical images, the labeled images and the unlabeled images are effectively utilized, the requirement for imagelabeling is lowered, the training data volume is increased, and therefore the prediction effect of the model can be improved while labeling resources are saved.

Description

technical field [0001] The present application relates to the field of artificial intelligence, and in particular to a method for training an image recognition model, a method and a device for image recognition. Background technique [0002] As the population continues to increase, the load on the medical system is increasing day by day, and the demand for medical resources is also increasing. In practical applications, medical staff can analyze the patient's condition through medical images. In order to help medical staff to diagnose diseases faster and more accurately, medical images can also be recognized with the help of automatic diagnostic equipment. [0003] At present, in the process of automatic diagnosis, a large number of medical images need to be used for training. Among them, these medical images need to be labeled by medical staff, that is, medical staff can make judgments on each medical image according to clinical habits, for example, label the medical image...

AI Technical Summary

Problems solved by technology

[0004] However, as the number of medical images continues to accumulate, the complexity of lesions becomes higher and higher, and the difficulty of labeling becomes more and more difficult.

However, due to limited annotation resources, only a small number of annotated medical images can be used in the model training process.

Moreover, since model training usually needs to be implemented in combination with specific tasks, different tasks need to use training sets corresponding to the tasks, resulting in the marked medical images not being effectively used and the data in the training sets of some tasks insufficient. The accuracy of the model's prediction effect is low

Images