基于广义高斯分布建模帧级变换系数的VVC编码失真预测模型

doi:10.11959/j.issn.1000-0801.2023088

Abstract

Abstract:

In versatile video coding (VVC), a variety of advanced coding tools work together to achieve excellent coding performance.Compared with high efficient video coding (HEVC), the transform coefficient distribution (TCD) of VVC has sharper peaks.In order to solve this phenomenon, the probability density function (PDF) of frame-level TCD was modeled, and a frame-level coding distortion prediction model based on statistical modeling was proposed, which modeled frame-level distortion as a function of TCD distribution parameters and quantization parameters.The experimental results show that compared with the Laplace distribution and Cauchy distribution, the generalized Gaussian distribution has the best performance in TCD probability density fitting.The prediction results based on the generalized Gaussian distribution distortion prediction model are closest to the actual coding distortion.

Key words: VVC, distortion estiamtion, D-Q model

CLC Number:

TP393

Yiyin GU, Hongkui WANG, Haibin YIN. VVC coded distortion prediction model based on frame-level transform coefficient modeling of generalized Gaussian distribution[J]. Telecommunications Science, 2023, 39(4): 101-110.

Figures/Tables 10

序列	QP		Lap			Cau			GGD
序列	QP	AI	LDP	RA	AI	LDP	RA	AI	LDP	RA
	22	0.170	0.040	0.226	0.160	0.043	0.189	0.161	0.038	0.181
BasketballDrive	27	0.113	0.075	0.037	0.093	0.084	0.087	0.099	0.133	0.038
	32	0.117	0.103	0.157	0.124	0.111	0.143	0.107	0.092	0.243
	37	0.128	0.109	0.270	0.125	0.098	0.164	0.123	0.100	0.233
	22	0.047	0.060	0.607	0.022	0.064	0.319	0.016	0.028	0.160
FourePeople	27	0.051	0.133	1.017	0.037	0.104	0.242	0.023	0.035	0.133
	32	0.038	0.067	0.608	0.041	0.063	0.343	0.028	0.046	0.160
	37	0.041	0.058	0.318	0.044	0.042	0.435	0.038	0.049	0.190
	22	0.070	0.255	0.325	0.063	0.166	0.049	0.059	0.049	0.199
BasketballDrill	27	0.072	0.117	0.213	0.059	0.071	0.048	0.066	0.055	0.131
	32	0.068	0.074	0.155	0.069	0.064	0.065	0.065	0.053	0.123
	37	0.066	0.061	0.134	0.077	0.049	0.099	0.064	0.056	0.117
	22	0.045	0.047	0.086	0.038	0.041	0.118	0.028	0.036	0.089
BQSquare	27	0.045	0.034	0.096	0.043	0.039	0.207	0.037	0.032	0.081
	32	0.033	0.041	0.117	0.033	0.049	0.197	0.031	0.038	0.122
	37	0.037	0.055	0.184	0.046	0.042	0.097	0.037	0.054	0.198
	22	0.029	0.021	0.031	0.013	0.011	0.027	0.008	0.008	0.029
Cactus	27	0.021	0.016	0.128	0.023	0.019	0.106	0.009	0.012	0.055
	32	0.021	0.029	0.034	0.022	0.010	0.049	0.016	0.023	0.074
	37	0.027	0.035	0.106	0.026	0.029	0.084	0.025	0.034	0.099
	22	0.067	0.078	0.828	0.026	0.071	0.753	0.015	0.016	0.138
KristenAndSara	27	0.045	0.101	1.192	0.033	0.105	0.913	0.020	0.024	0.124
	32	0.037	0.077	0.756	0.064	0.062	0.642	0.026	0.029	0.115
	37	0.034	0.050	0.401	0.023	0.033	0.315	0.028	0.029	0.146
	22	0.043	0.317	0.587	0.049	0.029	0.346	0.019	0.054	0.454
PartyScene	27	0.035	0.147	0.365	0.036	0.044	0.198	0.028	0.072	0.297
	32	0.035	0.087	0.246	0.033	0.049	0.169	0.033	0.066	0.217
	37	0.038	0.058	0.182	0.041	0.068	0.044	0.038	0.050	0.177
	22	0.051	0.054	0.370	0.039	0.072	0.243	0.032	0.043	0.367
BlowingBubbles	27	0.060	0.073	0.266	0.043	0.046	0.156	0.052	0.070	0.257
	32	0.073	0.091	0.220	0.063	0.011	0.098	0.071	0.090	0.204
	37	0.077	0.092	0.188	0.064	0.049	0.101	0.076	0.092	0.186
平均值		0.057	0.082 9	0.326	0.052	0.057	0.220	0.046	0.050	0.166

References 19

[1]	SULLIVAN G J , OHM J R , HAN W J ,et al. Overview of the high efficiency video coding (HEVC) standard[J]. IEEE Transactions on Circuits and Systems for Video Technology, 2012,22(12): 1649-1668.
[2]	BROSS B . Versatile video coding (draft 1):JVET-J1001[S]. 2018.
[3]	Qualcomm Inc. Rate distortion optimized quantization[EB]. 2008.
[4]	CHEN J , YE Y , KIM S . Algorithm description for versatile video coding and test model 10 (VTM 10.0)[EB]. 2020.
[5]	CHEN L N , YANG G B , HO A T . A Cauchy distribution based video watermark detection for H.264/AVC in DCT domain[C]// Proceedings of 2011 IEEE International Symposium of Circuits and Systems (ISCAS). Piscataway:IEEE Press, 2011: 2665-2668.
[6]	PANG C , AU O C , DAI J J ,et al. LMM-based frame-level rate control for H.264/AVC high-definition video coding[J]. Signal Processing:Image Communication, 2012,27(7): 737-748.
[7]	KANG J W . Novel distribution model of transformed coefficients in video coding using quad-tree structured block partitioning[J]. Multidimensional Systems and Signal Processing, 2017,28(4): 1589-1609.
[8]	SI J J , MA S W , WANG S Q ,et al. Laplace distribution based CTU level rate control for HEVC[C]// Proceedings of 2013 Visual Communications and Image Processing (VCIP). Piscataway:IEEE Press, 2014: 1-6.
[9]	HANG H M , CHEN J J . Source model for transform video coder and its application.I.Fundamental theory[J]. IEEE Transactions on Circuits and Systems for Video Technology, 1997,7(2): 287-298.
[10]	ALTUNBASAK Y , KAMACI N . An analysis of the DCT coefficient distribution with the H.264 video coder[C]// Proceedings of 2004 IEEE International Conference on Acoustics,Speech,and Signal Processing. Piscataway:IEEE Press, 2004.
[11]	KAMACI N , ALTUNBASAK Y , MERSEREAU R M . Frame bit allocation for the H.264/AVC video coder via Cauchy-density-based rate and distortion models[J]. IEEE Transactions on Circuits and Systems for Video Technology, 2005,15(8): 994-1006.
[12]	ZHAO X , SUN J , MA S W ,et al. Novel statistical modeling,analysis and implementation of rate-distortion estimation for H.264/AVC coders[J]. IEEE Transactions on Circuits and Systems for Video Technology, 2010,20(5): 647-660.
[13]	LI Y , MOU X Q . Transform coefficients distribution of the future versatile video coding (VVC) standard[C]// Proceedings of SPIE/COS Photonics Asia,SPIE 10817,Optoelectronic Imaging and Multimedia Technology V.[S.l.:s.n.], 2018: 120-130.
[14]	SUN J , DUAN Y Z , LI J T ,et al. Rate-distortion analysis of dead-zone plus uniform threshold scalar quantization and its application:part I:fundamental theory[J]. IEEE Transactions on Image Processing:a Publication of the IEEE Signal Processing Society, 2013,22(1): 202-214.
[15]	YANG E H , YU X , MENG J ,et al. Transparent composite model for DCT coefficients:design and analysis[J]. IEEE Transactions on Image Processing:a Publication of the IEEE Signal Processing Society, 2014,23(3): 1303-1316.
[16]	GAO W , KWONG S , YUAN H ,et al. DCT coefficient distribution modeling and quality dependency analysis based frame-level bit allocation for HEVC[J]. IEEE Transactions on Circuits and Systems for Video Technology, 2016,26(1): 139-153.
[17]	LEE B , KIM M . Modeling rates and distortions based on a mixture of Laplacian distributions for inter-predicted residues in quadtree coding of HEVC[J]. IEEE Signal Processing Letters, 2011,18(10): 571-574.
[18]	CUI J , WANG S S , WANG S Q ,et al. Hybrid Laplace distribution-based low complexity rate-distortion optimized quantization[J]. IEEE Transactions on Image Processing:a Publication of the IEEE Signal Processing Society, 2017,26(8): 3802-3816.
[19]	BROSS B , CHEN J , SHAN L Versatile video coding (draft 2):JVET-K1001[S]. 2018.

Metrics

Recommended 0

No Suggested Reading articles found!

序列	D_kl(Lap)			D_kl(Cau)			D_kl(GGD)
序列	I帧	P帧	B帧	I帧	P帧	B帧	I帧	P帧	B帧
Basketball	0.796	1.043	1.132	0.281	0.158	0.274	0.008	0.114	0.136
BlowingBubbles	0.122	0.434	0.637	0.104	0.035	0.092	0.126	0.113	0.188
BQMall	0.195	1.315	1.267	0.227	0.034	0.083	0.042	0.033	0.087
PeopleOnStreet	0.381	0.412	0.447	0.087	0.013	0.057	0.038	0.040	0.056
平均值	0.376 3	0.811	0.871	0.174	0.060	0.126	0.054	0.075	0.117

帧类型	Cau模型				GGD模型
帧类型	γ₁	γ₂	γ₃	γ₄	γ₁	γ₂	γ₃	γ₄
I帧	0.004 9	1.178 1	0.212 2	1.129 1	0.005 3	6.512 4	-0.555	0.837 4
P帧	0.002 2	1.049 1	0.274 2	0.854 7	0.001 6	2.897 4	-0.214	1.119 7
B帧	0.016 2	1.078 6	0.408 2	0.844 4	0.00 8	8.904 9	-0.31	0.831 7

权重系数	I帧	P帧	B帧
ω₁	0.533	0.412	0.359
ω₂	0.333	0.204	0.312
ω₃	0.144	0.384	0.329

VVC coded distortion prediction model based on frame-level transform coefficient modeling of generalized Gaussian distribution

RichHTML

PDF下载

Knowledge

Abstract

Cite this article

share this article

Figures/Tables 10

References 19

Related Articles 2

Metrics

Recommended 0

[1]	Wei QI, Haibing YIN, Hongkui WANG, Xiaofeng HUANG, Weihong NIU. Statistical modeling based fast rate estimation algorithm for VVC [J]. Telecommunications Science, 2022, 38(12): 35-45.
[2]	Shuang PENG, Xiaodong WANG, Zongju PENG, Fen CHEN. Fast QTMT partition decision based on deep learning [J]. Telecommunications Science, 2021, 37(4): 73-81.