US 12,335,657 B2
Method for processing video with dynamic video-based super-resolution network and circuit system
Kang-Yu Liu, Hsinchu (TW); Chon-Hou Sio, Hsinchu (TW); Chia-Wei Yu, Hsinchu (TW); and Hao-Ran Wang, Changsha (CN)
Assigned to REALTEK SEMICONDUCTOR CORP., Hsinchu (TW)
Filed by REALTEK SEMICONDUCTOR CORP., Hsinchu (TW)
Filed on Nov. 6, 2023, as Appl. No. 18/387,094.
Claims priority of application No. 202211392990.9 (CN), filed on Nov. 8, 2022.
Prior Publication US 2024/0155070 A1, May 9, 2024
Int. Cl. H04N 7/01 (2006.01); G06T 3/40 (2024.01); G06T 3/4053 (2024.01); G06T 7/00 (2017.01); G06V 10/44 (2022.01)
CPC H04N 7/0117 (2013.01) [G06T 3/4053 (2013.01); G06T 7/0002 (2013.01); G06V 10/44 (2022.01); G06T 2207/30168 (2013.01)] 16 Claims
OG exemplary drawing
 
1. A method for processing a video with a dynamic video-based super-resolution network, comprising:
receiving an input video;
continuously calculating, according to image features of the input video, quality scores of multiple frames of the input video, so as to assess a quality of the input video;
performing a moving average algorithm upon the quality scores of the multiple frames, so as to obtain a moving average score;
selecting, according to the moving average score, two corresponding weight tables for calculating a blending weight applied to a neural network super-resolution algorithm; and
applying the blending weight to the neural network super-resolution algorithm, so as to produce an output video.