Deep learning-Based Quality Enhancement Algorithms for Background of Video
Kei Kobayashi, Takafumi Katayama, Tian Song, Takashi Shimamoto · 2022
In this work, three quality enhancement algorithms are proposed to decrease the noise between the foreground and the background. The proposed algorithm is performed on the decoder side target to enhance the quality of the background with a higher quantization parameter to the same level of the foreground. The simulation results show that all three algorithms can improve the PSNR when using the proposed algorithms. The performance of these three algorithms is also discussed.