Image target detection model attack method and device, terminal equipment and storage medium

Abstract

The invention discloses an image target detection model attack method, and the method comprises the following steps: obtaining a sample image and disturbance image setting information inputted by a user, wherein the disturbance image setting information comprises the size and shape of an image; generating a disturbance image based on the sample image and the disturbance image setting information;generating an adversarial sample based on the sample image and the disturbance image; and inputting the confrontation sample into the image target detection model to be recognized so as to attack theimage target detection model. Since the size and the shape of the disturbance image are set by a user and the user setting has randomness and uncertainty, the difference between the confrontation sample and the sample image is not easy to perceive by naked eyes, the disturbance concealment is high, and the attack intensity is large. The invention further discloses an image target detection model attack device, terminal equipment and a storage medium, the concealment of disturbance is improved, and therefore the attack strength on the image target detection model is improved.

Description

technical field [0001] The invention relates to the field of image processing, in particular to an image target detection model attack method, device, terminal equipment and storage medium. Background technique [0002] With the development of machine learning, image recognition has been continuously improved, and the recognition accuracy has been continuously improved. [0003] However, adding adversarial perturbation to the image can make the image object detection model get completely different results. Therefore, in order to improve the robustness of the image object detection model, it is usually necessary to conduct adversarial attacks on the image object detection model through adversarial examples to improve the robustness of the image object detection model. Different adversarial samples have different attack effects. The stronger the attack, the more robust the image target detection model after the adversarial attack. Therefore, the selection of adversarial exam...

AI Technical Summary

Problems solved by technology

[0005] The main purpose of the present invention is to provide an image target detection model attack method, device, terminal equipment and storage medium, aiming to solve the technical problem of low attack strength of existing adversarial samples

Images