Deep learning model poisoning attack detection method and device based on mutual information

Abstract

The invention discloses a deep learning model poisoning attack detection method based on mutual information, comprising the following steps: (1) acquiring a sample set and a deep learning model to be detected; (2) pre-training the deep learning model; (3) generating a poisoning model Pool and trigger sample pair; (4) Use mutual information to determine whether the deep learning model is poisoned. The invention also discloses a deep learning model poisoning attack detection device based on mutual information, which is used to implement the above method. The invention has good applicability, can effectively judge whether the model is poisoned and find out the poisoning target class, and can achieve better detection effect.

Description

technical field [0001] The invention relates to the technical field of poisoning detection, in particular to a deep learning model poisoning attack detection method and device based on mutual information. Background technique [0002] Deep learning has gradually become a research hotspot and mainstream development direction in the field of artificial intelligence. Deep learning is a computational model consisting of multiple processing layers, learning machine learning techniques for data representation with multiple levels of abstraction. Deep learning represents the main development direction of machine learning and artificial intelligence research, bringing revolutionary progress to fields such as machine learning and computer vision. Artificial intelligence technology has made breakthroughs in the fields of computer vision and natural language processing, which has ushered in a new round of explosive development of artificial intelligence. Deep learning is the key to t...

AI Technical Summary

Problems solved by technology

Most existing testing techniques for autonomous driving rely on manual collection of test data for different driving conditions, which becomes unacceptably expensive as test scenarios increase

At the same time, the existing testing technologies are all based on the detection of poisoning attacks that are visible to triggers. The detection effect on feature embedding attacks that are not visible to triggers is very poor, and there are problems such as time-consuming and low efficiency in the detection process.

Images