Title:
RECOGNITION MODEL TRAINING METHOD AND APPARATUS, IMAGE RECOGNITION METHOD AND APPARATUS, AND DEVICE AND MEDIUM
Document Type and Number:
WIPO Patent Application WO/2021/017261
Kind Code:
A1
Abstract:
Disclosed are a recognition model training method and apparatus, an image recognition method and apparatus, and a device and a medium. The recognition model training method comprises: acquiring an original positive sample image and an original negative sample image carrying first labeling information; performing down-sampling processing on the original positive sample image and the original negative sample image, and performing screenshot processing by means of a screenshot tool to acquire a target positive sample image and a target negative sample image carrying second labeling information; inputting the target positive sample image and the target negative sample image into an MB-FCN model for model training so as to acquire a target MB-FCN detector; taking, as a group of target training data, the original positive sample image and the target positive sample image with a current image identifier in the first labeling information matching a source image identifier in the second labeling information; and inputting the target training data into a GAN model for model training so as to acquire a target GAN model. The target GAN model can help to improve the facial recognition accuracy of a blurred image.
Inventors:
WANG JIANZONG (CN)
PENG JUNQING (CN)
QU XIAOYANG (CN)
PENG JUNQING (CN)
QU XIAOYANG (CN)
Application Number:
PCT/CN2019/116489
Publication Date:
February 04, 2021
Filing Date:
November 08, 2019
Export Citation:
Assignee:
PING AN TECH SHENZHEN CO LTD (CN)
International Classes:
G06K9/00
Foreign References:
| CN108334848A | 2018-07-27 | |||
| CN108805809A | 2018-11-13 | |||
| US10325201B1 | 2019-06-18 |
Attorney, Agent or Firm:
SHENZHEN ZHONGDING INTELLECTUAL PROPERTY AGENCY (CN)
Download PDF: