Title:
METHOD AND APPARATUS FOR TRAINING FACIAL RECOGNITION MODEL, AND COMPUTER PROGRAM PRODUCT
Document Type and Number:
WIPO Patent Application WO/2023/016007
Kind Code:
A1
Abstract:
The present disclosure relates to the field of artificial intelligence, and particularly relates to computer vision and deep learning technology. Provided are a method and apparatus for training a facial recognition model, and an electronic device, a storage medium and a computer program product, which can be applied to a facial recognition scenario. The specific implementation solution involves: acquiring a training sample set, wherein training samples in the training sample set comprise sample facial images and category labels; and by using a machine learning method, taking the sample facial images as inputs and taking the category labels, which correspond to the input sample facial images, as expected outputs of two target fully connected layers in an initial facial recognition model, performing training to obtain a facial recognition model, wherein the two target fully connected layers sequentially model a sample facial image which comprises a covering object and a sample facial image which does not comprise a covering object. By means of the present disclosure, the recognition accuracy of a facial recognition model is improved.
Inventors:
LI BI (CN)
PENG NAN (CN)
XI TENG (CN)
ZHANG GANG (CN)
PENG NAN (CN)
XI TENG (CN)
ZHANG GANG (CN)
Application Number:
PCT/CN2022/092647
Publication Date:
February 16, 2023
Filing Date:
May 13, 2022
Export Citation:
Assignee:
BEIJING BAIDU NETCOM SCI & TECH CO LTD (CN)
International Classes:
G06K9/00
Foreign References:
| CN113657269A | 2021-11-16 | |||
| CN112052789A | 2020-12-08 | |||
| CN112767329A | 2021-05-07 | |||
| CN112734641A | 2021-04-30 | |||
| CN113221732A | 2021-08-06 | |||
| CN112115866A | 2020-12-22 | |||
| US20200042775A1 | 2020-02-06 |
Attorney, Agent or Firm:
INSIGHT INTELLECTUAL PROPERTY LIMITED (CN)
Download PDF: