Login| Sign Up| Help| Contact|

Patent Searching and Data


Title:
METHOD AND APPARATUS FOR CONVERTING VIDEO FRAME RATE ON BASIS OF DEEP LEARNING
Document Type and Number:
WIPO Patent Application WO/2020/105753
Kind Code:
A1
Abstract:
Provided are a video frame rate conversion method and apparatus for interpolating a frame with high quality at high speed while simultaneously responding to various color image formats on the basis of deep learning. A video frame rate conversion apparatus according to an embodiment of the present invention comprises: an adjustment unit for adjusting the resolution of an image of each channel; a generation unit for generating an intermediate image by using two consecutive t-second and (t+1)-second images of which the resolutions have been adjusted by the adjustment unit; and an enhancement unit for enhancing the resolution of the intermediate image generated by the generation unit. Therefore, when a deep learning-based video frame rate is converted, an input of the deep learning can be applied to an RGB color format and various color formats including YCbCr and the like, and high-quality and high-speed frame rate conversion is possible by simultaneously using a frame interpolation method and a resolution enhancement method.

Inventors:
JEONG JIN WOO (KR)
AHN HA EUN (KR)
KIM JE WOO (KR)
Application Number:
PCT/KR2018/014362
Publication Date:
May 28, 2020
Filing Date:
November 21, 2018
Export Citation:
Click for automatic bibliography generation   Help
Assignee:
KOREA ELECTRONICS TECHNOLOGY (KR)
International Classes:
H04N7/01; G06N3/08
Foreign References:
US20090279111A12009-11-12
KR20150004916A2015-01-13
JP2017041820A2017-02-23
US9288464B22016-03-15
Other References:
SIMON NIKLAUS: "Video Frame Interpolation via Adaptive Separable Convolution", IEEE INTERNATIONAL CONFERENCE ON COMPUTER VISION (ICCV, vol. 29, 22 October 2017 (2017-10-22), pages 261 - 270, XP033282880, DOI: 10.1109/ICCV.2017.37
Attorney, Agent or Firm:
NAM, Choong Woo (KR)
Download PDF: