METHOD AND APPARATUS FOR TRAINING IMAGE NOISE REDUCTION MODEL, ELECTRONIC DEVICE, AND STORAGE MEDIUM

WIPO Patent Application WO/2021/253316

A1

Disclosed in the embodiments of the present description are a method and apparatus for training an image noise reduction model, an electronic device, and a computer-readable storage medium. The method comprises: inputting an original noise-containing image into a first generative model of a cycle-consistent generative adversarial network, and generating a first noise-free generative image; inputting the first noise-free generative image into a second generative model of the cycle-consistent generative adversarial network, and generating a second noise-containing generative image; inputting a noise-free original image into the second generative model of the cycle-consistent generative adversarial network, and generating a first noise-containing generative image; inputting the first noise-containing generative image into the first generative model of the cycle-consistent generative adversarial network, and generating a second noise-free generative image; and optimizing the cycle-consistent generative adversarial network so as to obtain a trained image noise reduction model.

ZHENG HAIRONG (CN)
LIU XIN (CN)
ZHANG NA (CN)
HU ZHANLI (CN)
XUE HENGZHI (CN)
LIANG DONG (CN)

PCT/CN2020/096770

December 23, 2021

June 18, 2020

Click for automatic bibliography generation  

SHENZHEN INST ADV TECH (CN)

G06T5/50

CN110390647A	2019-10-29
CN110634108A	2019-12-31
CN111275647A	2020-06-12
US20200111194A1	2020-04-09

COHORIZON INTELLECTUAL PROPERTY INC. (CN)

View/Download PDF      PDF Help

Previous Patent: DYNAMIC RANDOM ACCESS MEMORY APPARATUS

Next Patent: ELECTRICAL PROTECTIVE DEVICE FOR LOW-VOLTAGE DIRECT CURRENT (LVDC) NETWORK