Login| Sign Up| Help| Contact|

Patent Searching and Data


Title:
ELECTRONIC DEVICE FOR PERFORMING INFERENCE ON BASIS OF ENCRYPTED INFORMATION BY USING ARTIFICIAL INTELLIGENCE MODEL, AND OPERATING METHOD THEREFOR
Document Type and Number:
WIPO Patent Application WO/2022/158686
Kind Code:
A1
Abstract:
Provided is a method by which an electronic device performs an inference operation on the basis of encrypted information, received from an external device, by using an artificial intelligence model, the method comprising: converting the artificial intelligence model so that the artificial intelligence model can perform inference on the basis of encrypted information; generating, on the basis of the converted artificial intelligence model, parameter information including information related to at least one parameter for encrypting information to be input into the artificial intelligence model; transmitting the parameter information to the external device; receiving information encrypted on the basis of the parameter information from the external device; and inputting the encrypted information into the converted artificial intelligence model to acquire an inference result output from the converted artificial intelligence model.

Inventors:
KIM JUNGKON (KR)
KIM JINSU (KR)
SHIN JUNBUM (KR)
Application Number:
PCT/KR2021/015447
Publication Date:
July 28, 2022
Filing Date:
October 29, 2021
Export Citation:
Click for automatic bibliography generation   Help
Assignee:
SAMSUNG ELECTRONICS CO LTD (KR)
International Classes:
G06N5/02; G06N3/08; G06N20/00
Domestic Patent References:
WO2018151552A12018-08-23
Foreign References:
US20160350648A12016-12-01
US20170317983A12017-11-02
Other References:
EDWARD CHOU; JOSH BEAL; DANIEL LEVY; SERENA YEUNG; ALBERT HAQUE; LI FEI-FEI: "Faster CryptoNets: Leveraging Sparsity for Real-World Encrypted Inference", ARXIV.ORG, CORNELL UNIVERSITY LIBRARY, 201 OLIN LIBRARY CORNELL UNIVERSITY ITHACA, NY 14853, 25 November 2018 (2018-11-25), 201 Olin Library Cornell University Ithaca, NY 14853 , XP080938566
XU RUNHUA; JOSHI JAMES B.D.; LI CHAO: "CryptoNN: Training Neural Networks over Encrypted Data", 2019 IEEE 39TH INTERNATIONAL CONFERENCE ON DISTRIBUTED COMPUTING SYSTEMS (ICDCS), IEEE, 7 July 2019 (2019-07-07), pages 1199 - 1209, XP033651771, DOI: 10.1109/ICDCS.2019.00121
Attorney, Agent or Firm:
Y.P.LEE, MOCK & PARTNERS (KR)
Download PDF: