Method for protecting embedded machine learning model

The invention relates to a method (100) for protecting an embedded machine learning model (200) against at least one physical attack, comprising:-determining (101) a monitoring input (310), where the monitoring input (310) is based on at least one intermediate result (210) of the machine learning mo...

Full description

Saved in:
Bibliographic Details
Main Authors HETWELL BRIAN, SCHORN CHRISTIAN
Format Patent
LanguageChinese
English
Published 11.06.2024
Subjects
CALCULATING
COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS
COMPUTING
COUNTING
ELECTRIC COMMUNICATION TECHNIQUE
ELECTRICITY
PHYSICS
TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHICCOMMUNICATION
Online AccessGet full text

Cover

More Information
Summary:The invention relates to a method (100) for protecting an embedded machine learning model (200) against at least one physical attack, comprising:-determining (101) a monitoring input (310), where the monitoring input (310) is based on at least one intermediate result (210) of the machine learning model (200),-evaluating (102) the determined monitoring input (310) by a monitoring system (300), -detecting (103) at least one physical attack based on the evaluation (102). 本发明涉及一种用于针对至少一种物理攻击来保护嵌入式机器学习模型(200)的方法(100)，所述方法包括：-确定(101)监控输入(310)，其中所述监控输入(310)基于所述机器学习模型(200)的至少一个中间结果(210)，-通过监控系统(300)评估(102)所确定的监控输入(310)，-基于所述评估(102)而检测(103)至少一种物理攻击。
Bibliography:Application Number: CN202311691317