Adversarial sample and generation method thereof, medium, device and computing equipment

Abstract

The embodiment of the invention provides an adversarial sample generation method. The method comprises the following steps: respectively acquiring an image of an attack object and an image of an attacked object; selecting a feature image to be fused according to the image of the attacked object; performing image fusion based on the image of the attack object and the feature image; and obtaining anadversarial sample based on the fused image. According to the method disclosed by the invention, the adversarial sample is generated; a model which is the same as or similar to the attacked model does not need to be acquired and / or operated; in addition, according to one method disclosed by the invention, an adversarial sample based on image fusion can be adopted as initialization, the attack success rate of an existing adversarial sample generation method based on optimization can be remarkably improved, and the embodiment of the invention provides an adversarial sample generation device, amedium and computing equipment.

Description

technical field [0001] Embodiments of the present invention relate to the technical field of computer vision, and more specifically, embodiments of the present invention relate to an adversarial example and its generation method, medium, device, and computing device. Background technique [0002] This section is intended to provide a background or context for implementations of the invention that are recited in the claims. The descriptions herein are not admitted to be prior art by inclusion in this section. [0003] Adversarial example generation methods represent a class of methods for attacking machine learning models during the testing phase. Existing adversarial example generation methods can be divided into optimization-based methods and non-optimization-based methods. [0004] Among optimization-based methods, different adversarial example generation methods differ mainly in the optimization algorithm and perturbation form. For common white-box-based attacks and bl...

AI Technical Summary

Problems solved by technology

It is obvious that optimization-based adversarial example generation methods need to obtain or run a model that is the same or similar to the victim model, and this model is often not easy to obtain and requires high computing resources to run

[0005] Although the existing non-optimization-based adversarial sample generation methods do not need to obtain or run a model that is the same as or similar to the victim model, they often only use simple geometric patterns or image transformations to impose disturbances, and it is difficult to attack complex semantic information. Machine learning models (such as face comparison and object detection)

Images