SECURING SYSTEMS EMPLOYING ARTIFICIAL INTELLIGENCE

Techniques and apparatuses to harden AI systems against various attacks are provided. Among the different techniques and apparatuses, is provided, techniques and apparatuses that expand the domain for an inference model to include both visible classes and well as hidden classes. The hidden classes c...

Full description

Saved in:
Bibliographic Details
Main Authors KELLERMANN, Raizy, NAYSHTUT, Alex, SUKHOMLINOV, Vadim, BEN-SHALOM, Omer, BARNHART-MAGEN, Guy, POGORELIK, Oleg, KLIMOV, Denis
Format Patent
LanguageEnglish
French
Published 09.07.2020
Subjects
CALCULATING
COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS
COMPUTING
COUNTING
PHYSICS
Online AccessGet full text

Cover

More Information
Summary:Techniques and apparatuses to harden AI systems against various attacks are provided. Among the different techniques and apparatuses, is provided, techniques and apparatuses that expand the domain for an inference model to include both visible classes and well as hidden classes. The hidden classes can be used to detect possible probing attacks against the model. L'invention concerne des techniques et des appareils permettant de durcir des systèmes d'intelligence artificielle contre diverses attaques. Parmi les différents techniques et appareils, l'invention concerne des techniques et des appareils qui étendent le domaine pour un modèle d'inférence afin d'inclure à la fois des classes visibles et des classes cachées. Les classes cachées peuvent être utilisées pour détecter de possibles attaques par sondage contre le modèle.
Bibliography:Application Number: WO2019US28687