Target detection model training method, device, electronic equipment and storage medium

Abstract

The invention relates to the technical field of artificial intelligence, and can be applied to a smart community scene, thereby promoting the construction of a smart city. The target detection model training method comprises the steps of acquiring a preset training set, wherein each training image of the training set carries a detection box used for marking an object; generating a plurality of candidate boxes in each detection box according to the preset coordinate increment and the preset proportion; determining the gradient sum of each candidate box; determining a target candidate box from aplurality of candidate boxes in each detection box according to the gradient sum; performing norm disturbance on image pixels of the target candidate frame, so that an adversarial patch image can beobtained; and optimizing the target detection model in training by using the adversarial patch image to obtain a trained target detection model. The invention further provides a target detection modeltraining device, electronic equipment and a storage medium. According to the invention, the identification accuracy of the target detection model can be improved.

Description

technical field [0001] The present invention relates to the technical field of artificial intelligence, in particular to a target detection model training method, device, electronic equipment and storage medium. Background technique [0002] At present, with the development of artificial intelligence technology, different objects in the image can be detected by the target detection model obtained by machine learning, neural network training, etc., but in practice, it is found that the target detection model is vulnerable to adversarial attacks. For example, the detection object wears a mask, wears glasses, or is partially covered by other objects. At this time, the target detection model is prone to errors, and the recognition accuracy is not high. [0003] Therefore, how to improve the recognition accuracy of the target detection model is a technical problem that needs to be solved urgently. Contents of the invention [0004] In view of the above, it is necessary to prov...

AI Technical Summary

Problems solved by technology

[0002] At present, with the development of artificial intelligence technology, different objects in the image can be detected by the target detection model obtained by machine learning, neural network training, etc., but in practice, it is found that the target detection model is vulnerable to adversarial attacks. For example, the detection object wears a mask, wears glasses, or is partially covered by other objects. At this time, the target detection model is prone to errors, and the recognition accuracy is not high.

Images