Patent · US Active

Method of training image deep learning model and device thereof

US11720790B2 · kind B2 · utility

0Cited by
0References
18Claims
0Family size

Assignees

Inventors

Key dates

Filing dateMay 21, 2020
Grant dateAug 8, 2023
Priority date
Expiry dateApr 20, 2041

Classification

  • Technology area (CPC G)Physics
  • CPC primaryG06V10/774
  • WIPO fieldComputer technology
  • WIPO sectorElectrical engineering

Abstract

Disclosed herein is an image deep learning model training method. The method includes sampling a twin negative comprising a first negative sample and a second negative sample by selecting the first negative sample with a highest similarity out of an anchor sample and a positive sample constituting a matching pair in each class and by selecting the second negative sample with a highest similarity to the first negative sample, and training the samples to minimize a loss of a loss function in each class by utilizing the anchor sample, the positive sample, the first and second negative samples for each class. The first negative sample is selected in a different class from a class comprising the matching pair, and the second negative sample is selected in a different class from classes comprising the matching pair and the first negative sample.

Source: USPTO / EPO open patent data. Objective bibliographic and citation counts.