Patent · US Active

Device based on machine learning

US12086956B2 · kind B2 · utility

1Cited by

3References

7Claims

0Family size

Assignee

VINAI AI APPLICATION AND RESEARCH JOINT STOCK CO. · VN

Inventors

Cuong PHAM · Harold, US
Tu Pham · San Jose, US
Dung Tri Nguyen · Hue City, VN

Key dates

Filing date	Feb 9, 2022
Grant date	Sep 10, 2024
Priority date	—
Expiry date	Mar 12, 2043

Classification

Technology area (CPC G)Physics
CPC primaryG06T2207/30196
WIPO fieldComputer technology
WIPO sectorElectrical engineering

Abstract

An image processing device includes a convolutional neural network (CNN) architecture configured to perform at least one of denoising, deblurring, dehazing, and color correction on an input. The input comprises a plurality of low-quality input images having different exposure values captured for the same object at the same time by a camera under display (CUD) to render a high-quality output image. The CNN architecture is trained using a set of training images including a plurality of low-quality training images having different exposure values and a high-quality training image corresponding to the plurality of low-quality training images.

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