TY - GEN
T1 - A learning-based low complexity in-loop filter for video coding
AU - Liu, Chao
AU - Sun, Heming
AU - Katto, Jiro
AU - Zeng, Xiaoyang
AU - Fan, Yibo
N1 - Funding Information:
This work was supported in part by the National Natural Science Foundation of China under Grant 61674041, in part by Alibaba Group through Alibaba Innovative Research (AIR) Program, in part by the STCSM under Grant 16XD1400300, in part by the pioneering project of academy for engineering and technology and Fudan-CIOMP joint fund, in part by the National Natural Science Foundation of China under Grant 61525401the Program of Shanghai Academic/Technology Research Leader under Grant 16XD1400300, the Innovation Program of Shanghai Municipal Education Commission, in part by JST, PRESTO Grant Number JPMJPR19M5, Japan.
Publisher Copyright:
© 2020 IEEE.
PY - 2020/7
Y1 - 2020/7
N2 - With the continuous development of mobile devices, it becomes possible for people to demand higher definition videos. To alleviate the pressure of deploying the video codec in mobile multimedia, a learning-based low complexity in-loop filter is proposed in this paper. Depthwise separable convolution is combined with batch normalization to construct this model. To enhance its performance, the knowledge from a pre-trained teacher model is transferred to it. However, the over-smoothing problem in the inter frames caused by double enhancing effect remains. To solve this, a Wiener-based filtering algorithm that tries to restore the distortion from the learned residuals is designed and introduces an adequate filtering effect. The experimental results show that our proposed methods achieve considerable BD-rate reduction than HEVC anchor. Compared with the previous state-of-the-art work VR-CNN, our model achieves 1.65% extra BD-rate reduction, 79.1% decrease in FLOPs, 25% decrease in encoding complexity, and 70% decoding complexity decrease.
AB - With the continuous development of mobile devices, it becomes possible for people to demand higher definition videos. To alleviate the pressure of deploying the video codec in mobile multimedia, a learning-based low complexity in-loop filter is proposed in this paper. Depthwise separable convolution is combined with batch normalization to construct this model. To enhance its performance, the knowledge from a pre-trained teacher model is transferred to it. However, the over-smoothing problem in the inter frames caused by double enhancing effect remains. To solve this, a Wiener-based filtering algorithm that tries to restore the distortion from the learned residuals is designed and introduces an adequate filtering effect. The experimental results show that our proposed methods achieve considerable BD-rate reduction than HEVC anchor. Compared with the previous state-of-the-art work VR-CNN, our model achieves 1.65% extra BD-rate reduction, 79.1% decrease in FLOPs, 25% decrease in encoding complexity, and 70% decoding complexity decrease.
KW - CNN
KW - HEVC
KW - In-loop filter
KW - Inter
KW - Knowledge transfer
UR - http://www.scopus.com/inward/record.url?scp=85091757857&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85091757857&partnerID=8YFLogxK
U2 - 10.1109/ICMEW46912.2020.9106015
DO - 10.1109/ICMEW46912.2020.9106015
M3 - Conference contribution
AN - SCOPUS:85091757857
T3 - 2020 IEEE International Conference on Multimedia and Expo Workshops, ICMEW 2020
BT - 2020 IEEE International Conference on Multimedia and Expo Workshops, ICMEW 2020
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 2020 IEEE International Conference on Multimedia and Expo Workshops, ICMEW 2020
Y2 - 6 July 2020 through 10 July 2020
ER -