Authentication method and apparatus with transformation model

Patent metadata

Field	Value
Publication number	US-12293765-B2
Application number	US-202318317503-A
Country	US
Kind code	B2
Filing date	May 15, 2023
Priority date	Mar 15, 2019
Publication date	May 6, 2025
Grant date	May 6, 2025

Abstract

An authentication method and apparatus using a transformation model are disclosed. The authentication method includes generating, at a first apparatus, a first enrolled feature based on a first feature extractor, obtaining a second enrolled feature to which the first enrolled feature is transformed, determining an input feature by extracting a feature from input data with a second feature extractor different from the first feature extractor, and performing an authentication based on the second enrolled feature and the input feature.

First claim

What is claimed is:

1. A processor implemented authentication method, the method comprising:
obtaining a transformation model corresponding to a difference between a first feature extractor and a second feature extractor, the first feature extractor comprising a first neural network and the second feature extractor comprising a second neural network separate from the first neural network; and
transforming an initial enrolled feature having been generated by the first feature extractor to an updated enrolled feature by using the transformation model,
wherein the updated enrolled feature is used for a user authentication in response to an input feature determined by extracting a feature from an authentication request including input authentication data with the second feature extractor.

2. The method of claim 1 , wherein the updated enrolled feature is provided an apparatus other than the apparatus that performs the transforming.

3. The method of claim 1 , wherein the second feature extractor is an updated version of the first feature extractor.

4. The method of claim 1 , wherein the transformation model includes a structural element that corresponds to a structural difference between the first feature extractor and the second feature extractor.

5. The method of claim 1 , wherein the first feature extractor is pretrained to output first features for use with a first authentication procedure, wherein the second feature extractor is pretrained to output second features for a second authentication procedure.

6. The method of claim 5 , wherein the transformation model is a neural network pretrained to transform features for use with the first authentication procedure to features for use with the second authentication procedure.

7. The method of claim 1 , wherein the initial enrolled feature includes initial sub-features, and the updated enrolled feature includes updated sub-features to which the initial sub-features are respectively transformed.

8. The method of claim 7 , further comprising:
discarding at least a portion of the updated sub features based on similarity measures of the updated sub-features with respect to each other.

9. The method of claim 8 , wherein the discarding comprises discarding at least one of the updated sub-features based on a similarity measure between the at least one of the updated sub-features and an aggregate similarity of other of the updated sub features.

10. The method of claim 9 , wherein the user authentication is based on a first threshold and a similarity between the updated enrolled feature and the input feature, and
wherein the first threshold is equal to a second threshold based upon which the at least one of the updated sub-features is discarded.

11. An authentication apparatus, comprising:
one or more processors configured to:
obtain a transformation model corresponding to a difference between a first feature extractor and a second feature extractor, the first feature extractor comprising a first neural network and the second feature extractor comprising a second neural network separate from the first neural network; and
transform an initial enrolled feature having been generated by the first feature extractor to an updated enrolled feature by using the transformation model,

wherein the updated enrolled feature is used for a user authentication in response to an input feature determined by extracting a feature from an authentication request including input authentication data with the second feature extractor.

12. The apparatus of claim 11 , wherein the second feature extractor is an updated version of the first feature extractor.

13. The apparatus of claim 11 , wherein the transformation model includes a structural element that is not a structural element of the first feature extractor and is a structural element of the second feature extractor.

14. The apparatus of claim 11 , wherein the first feature extractor is pretrained to output first features for use with a first authentication procedure, wherein the second feature extractor is pretrained to output second features for a second authentication procedure, and wherein the transformation model is a neural network pretrained to transform features for use with the first authentication procedure to features for use with the second authentication procedure.

15. The apparatus of claim 11 , wherein the initial enrolled feature includes initial sub-features, and the updated enrolled feature includes updated sub-features to which the initial sub-features are transformed.

16. The apparatus of claim 15 , wherein the one or more processors are configured to discard at least a portion of the updated sub-features based on a similarity between the updated sub-features.

17. The apparatus of claim 16 , wherein the one or more processors are configured to discard one of the updated sub-features based on a threshold and similarities between the one updated sub-feature and the remaining updated sub-features.

18. An IoT (internet of things) device, comprising
one or more processors configured to:
obtain a transformation model corresponding to a difference between a first feature extractor and a second feature extractor, the first feature extractor comprising a first neural network and the second feature extractor comprising a second neural network separate from the first neural network; and
transform an initial enrolled feature having been generated by the first feature extractor to an updated enrolled feature by using the transformation model,

wherein the updated enrolled feature is used for a user authentication in response to an input feature determined by extracting a feature from an authentication request including input authentication data with the second feature extractor.

19. The IoT device of claim 18 , wherein the updated enrolled feature is provided to another IoT device performing the user authentication.

Assignees

• Samsung Electronics Co Ltd
• Snu R&Db Foundation

Inventors

• Han Seungju
• Han Jaejoon
• Ko Minsu
• Choi Chang Kyu
• Han Bohyung

Classifications

• G06N3/0464

  Convolutional networks [CNN, ConvNet] · CPC title

• G06N3/082

  modifying the architecture, e.g. adding, deleting or silencing nodes or connections · CPC title

• G06N3/09

  Supervised learning · CPC title

• G06F18/20

  Analysing · CPC title

• G06V10/764

  using classification, e.g. of video objects · CPC title