基于SIMD的AVS3并行率失真优化量化算法

doi:10.11959/j.issn.1000-0801.2024155

Abstract

Abstract:

To improve the coding efficiency of rate-distortion optimization quantization (RDOQ) for the third-generation audio video coding standard (AVS3), a parallelized RDOQ algorithm based on single instruction multiple data (SIMD) was proposed. Firstly, in the optimal coefficient decision (OCD) stage, the dependencies within the scan line were retained through optimization. Secondly, in the last non-zero position decision (LNPD) stage, the Zig-Zag scan line was split into various independent lines based on the partitioning strategy. The optimal coefficients on each scan line were allowed to be calculated in parallel by the proposed method. Finally, SIMD instructions were utilized for parallelized acceleration to boost calculation efficiency in the RDOQ process. Experimental results show that the proposed algorithm can achieve a 29.46% coding time reduction with only 0.25% BD-Rate loss for AI configuration.

Key words: AVS3, RDOQ, SIMD, parallelized acceleration

CLC Number:

TN919.81

Yixin TANG, Xiaofeng HUANG, Ran TANG, Yang ZHOU, Yan CUI, Haibing YIN. SIMD-based parallelized rate distortion optimized quantization for AVS3[J]. Telecommunications Science, 2024, 40(6): 114-126.

Figures/Tables 16

\|l_i \|	候选值
0	0
2	0，1，2
$⋮$	$⋮$
N	0，N，N-1

References 29

1	司武旭. 5G通信技术传输超高清视频的相关研究[J]. 现代工业经济和信息化, 2022, 12(12): 54-55.
	SI W X. Research related to the transmission of ultra-high definition video by 5G communication technology[J]. Modern Industrial Economy and Informationization, 2022, 12(12): 54-55.
2	CHOI K, CHEN J L, RUSANOVSKYY D, et al. An overview of the MPEG-5 essential video coding standard in a nutshell[J]. IEEE Signal Processing Magazine, 2020, 37(3): 160-167.
3	ZHANG J, JIA C, LEI M, et al. Recent development of AVS video coding standard: AVS3[C]//Proceedings of 2019 picture coding symposium (PCS). Piscataway: IEEE Press, 2019: 1-5.
4	HUANG Y W, HSU C W, CHEN C Y, et al. A VVC proposal with quaternary tree plus binary-ternary tree coding block structure and advanced coding techniques[J]. IEEE Transactions on Circuits and Systems for Video Technology, 2020, 30(5): 1311-1325.
5	FU Y B, WANG S, ZHU C, et al. An attention based CNN with temporal hierarchical deployment for AVS3 inter In-loop filtering[C]//Proceedings of the 2022 IEEE International Symposium on Circuits and Systems (ISCAS). Piscataway: IEEE Press, 2022: 501-505.
6	姚英彪, 李晓娟. 基于图像空间相关性与纹理的HEVC块划分快速算法[J]. 电信科学, 2015, 31(1): 38-46.
	YAO Y B, LI X J. Fast block partitioning algorithm for HEVC based on spatial correlation and image texture[J]. Telecommunications Science, 2015, 31(1): 38-46.
7	祁伟, 殷海兵, 王鸿奎, 等. 基于统计建模的VVC快速码率估计算法[J]. 电信科学, 2022, 38(12): 35-45.
	QI W, YIN H B, WANG H K, et al. Statistical modeling based fast rate estimation algorithm for VVC[J]. Telecommunications Science, 2022, 38(12): 35-45.
8	孟翔, 殷海兵, 黄晓峰. 基于统计建模的HEVC快速率失真估计算法[J]. 电信科学, 2021, 37(1): 58-68.
	MENG X, YIN H B, HUANG X F. Statistical modeling based fast rate distortion estimation algorithm for HEVC[J]. Telecommunications Science, 2021, 37(1): 58-68.
9	ZHANG D D, CHEN Y W, IZQUIERDO E. Fast intra mode decision for HEVC based on texture characteristic from RMD and MPM[C]//Proceedings of the 2014 IEEE Visual Communications and Image Processing Conference. Piscataway: IEEE Press, 2014: 510-513.
10	WU Q B, XIONG J, LUO B, et al. A novel joint rate distortion optimization scheme for intra prediction coding in H.264/AVC[J]. IEICE Transactions on Information and Systems, 2014, 97(4): 989-992.
11	HE D K, WANG J. Rate distortion optimized quantization based on weighted mean squared error for lossy image coding[C]//Proceedings of the 2014 IEEE International Conference on Image Processing (ICIP). Piscataway: IEEE Press, 2014: 5606-5610.
12	DENG Y, PENG Q, YANG T W. A rate-distortion optimized real time intra-update method for packet video transmission[J]. Proceedings of the 2003 IEEE International Conference on Intelligent Transportation Systems, 2003(2): 915-918vol.2.
13	LEMMETTI A, KALLIO E, VIITANEN M, et al. Rate-distortion-complexity optimized coding scheme for kvazaar HEVC intra encoder[C]//Proceedings of the 2018 Data Compression Conference. Piscataway: IEEE Press, 2018: 419.
14	XU M T, NGUYEN CANH T, JEON B. Simplified level estimation for rate-distortion optimized quantization of HEVC[J]. IEEE Transactions on Broadcasting, 2020, 66(1): 88-99.
15	WANG M, XIE X D, FAN H F, et al. Fast rate distortion optimized quantization method for HEVC[C]//Proceedings of the 2017 IEEE International Symposium on Circuits and Systems (ISCAS). Piscataway: IEEE Press, 2017: 1-4.
16	WANG H, YIN H, SHEN Y. A novel hard-decision quantization algorithm based on adaptive deadzone offset model[C]//Advances in Multimedia Information Processing-PCM 2016: 17th Pacific-Rim Conference on Multimedia, Xi′an, China, September 15-16, 2016 , Proceedings, Part Ⅱ. Springer International Publishing, 2016: 335-345.
17	WANG J, YIN H B, GAO Z Y, et al. Improved rate distortion optimized quantization for HEVC with adaptive thresholding[C]// Proceedings of the 2016 IEEE International Symposium on Broadband Multimedia Systems and Broadcasting (BMSB). Piscataway: IEEE Press, 2016: 1-4.
18	LEE H, YANG S, PARK Y, et al. Fast quantization method with simplified rate–distortion optimized quantization for an HEVC encoder[J]. IEEE Transactions on Circuits and Systems for Video Technology, 2016, 26(1): 107-116.
19	YIN H B, CAI H, LU H J. A new all-zero block detection algorithm for high efficiency video coding[C]//Proceedings of the 2017 Data Compression Conference (DCC). Piscataway: IEEE Press, 2017: 470.
20	YANG H, YIN H, HUANG X. A fast zero-quantized percentage model for video coding with RDO quantization[C]//Advances in Multimedia Information Processing–PCM 2018: 19th Pacific-Rim Conference on Multimedia, Hefei, China, September 21-22, 2018, Proceedings, Part Ⅲ 19. Springer International Publishing, 2018: 761-771.
21	YIN H B, YANG E H, YU X, et al. Fast soft decision quantization with adaptive preselection and dynamic trellis graph[J]. IEEE Transactions on Circuits and Systems for Video Technology, 2015, 25(8): 1362-1375.
22	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.
23	HE J, YANG F Z. High-speed implementation of rate-distortion optimized quantization for H.264/AVC[J]. Signal, Image and Video Processing, 2015, 9(3): 543-551.
24	CANH T N, XU M, JEON B. Rate-distortion optimized quantization: A deep learning approach[C]//Proceedings of IEEE High Performance Extreme Computing Conference. Piscataway: IEEE Press, 2018.
25	IGARASHI H, TAKANO F, TAKENAKA T, et al. Parallel rate distortion optimized quantization for 4K real-time GPU-based HEVC encoder[C]//Proceedings of the 2018 IEEE Visual Communications and Image Processing (VCIP). Piscataway: IEEE Press, 2018: 1-4.
26	ZHAO J C, YANG F, HUANG X F, et al. Scanline-based fast algorithm and pipelined hardware design of rate-distortion optimized quantization for AVS3[C]//Proceedings of the 2023 IEEE International Conference on Consumer Electronics (ICCE). Piscataway: IEEE Press, 2023: 1-6.
27	CHEN K J, DUAN Y Z, YAN L J, et al. Efficient SIMD optimization of HEVC encoder over X86 processors[C]//Proceedings of the Proceedings of The 2012 Asia Pacific Signal and Information Processing Association Annual Summit and Conference. Piscataway: IEEE Press, 2012: 1-4.
28	BOSSEN F. Common test conditions and software reference configurations[J]. JCTVC-L1100, 2013, 12(7): 1.
29	BJONTEGAARD G. Calculation of average PSNR differences between RD-curves[J]. ITU SG16 Doc. VCEG-M33, 2001.

Metrics

Recommended 0

No Suggested Reading articles found!

函数指令	指令作用描述
__m128i_mm_setr_epi32	创建一个包含4个32位整数的SSE128寄存器
__m128i_mm_mullo_epi32	将a, b中对应的4个32位整数进行乘法运算
__m128i_mm_min_epi32	将a, b中对应的4个32位整数进行比较，选取每对整数中较小的输出
__m128i_mm_add_epi32	将a, b中对应的4个32位整数进行加法运算
__m128i _mm_blendv_epi8	根据mask的值选择a, b中的相应元素进行混合并返回结果

类	序列	AI			RA			LDB
类	序列	BDBR	TS_Q	TS_Enc	BDBR	TS_Q	TS_Enc	BDBR	TS_Q	TS_Enc
UHD	Tango2	0.16%	32.36%	9.81%	0.26%	30.56%	7.55%	0.28%	29.92%	8.00%
	Parkrunning3	0.09%	31.12%	8.10%	0.19%	26.04%	6.09%	0.19%	27.64%	7.07%
	Campfire	0.13%	30.42%	8.49%	0.35%	28.21%	7.42%	0.36%	28.32%	6.82%
	DaylightRoad2	0.08%	31.49%	8.31%	0.22%	28.55%	7.25%	0.17%	28.95%	6.95%
	Average	0.12%	31.35%	8.68%	0.26%	28.34%	7.08%	0.25%	28.71%	7.21%
HD	Cactus	0.35%	30.75%	8.22%	0.24%	26.03%	6.47%	0.30%	28.21%	6.65%
	BasketballDrive	0.46%	31.27%	8.32%	0.23%	27.52%	6.51%	0.34%	27.38%	6.45%
	MarketPlace	0.16%	32.01%	10.26%	0.44%	29.69%	7.23%	0.20%	31.04%	7.40%
	RitualDance	0.15%	31.29%	8.24%	0.33%	28.56%	7.63%	0.10%	28.89%	7.10%
	Average	0.28%	31.33%	8.76%	0.31%	27.95%	6.96%	0.24%	28.88%	6.90%
720p	City	0.36%	32.36%	10.00%	0.45%	29.55%	7.30%	0.49%	30.10%	8.12%
	Crew	0.34%	31.52%	8.56%	-0.29	29.08	6.50	0.42%	27.17%	6.40%
	vidyo1	0.15%	31.44%	8.20%	0.33%	28.15%	6.84%	0.20%	26.35%	6.32%
	vidyo3	0.25%	28.52%	8.00%	0.17%	27.56%	6.34%	-0.04%	28.64%	6.86%
	Average	0.28%	30.96%	8.69%	0.17%	28.59%	6.75%	0.27%	28.07%	6.93%
240p	BasketballPass	0.24%	30.59%	8.65%	0.14%	27.84%	6.88%	0.56%	28.42%	6.03%
	BQSquare	0.12%	32.56%	10.51%	0.53%	30.33%	7.08%	0.04%	28.85%	6.82%
	BlowingBubbles	0.36%	29.75%	7.84%	0.24%	27.13%	6.30%	0.41%	27.14%	6.67%
	RaceHorses	0.28%	34.21%	9.58%	0.09%	30.52%	7.99%	0.23%	29.96%	7.88%
	Average	0.25%	31.78%	9.15%	0.25%	28.96%	7.06%	0.31%	28.59%	6.85%
Summary	Overall	0.23%	31.35%	8.82%	0.25%	28.46%	6.96%	0.27%	28.56%	6.97%

方法	BD-Rate
方法	AI	RA	LDB
OCD并行优化	0.18%	0.21%	0.21%
OCD并行优化+基于分治策略的LNPD	0.23%	0.25%	0.27%

现有工作	参考软件	AI		RA		LDB
现有工作	参考软件	BD-Rate	TS_Q	BD-Rate	TS_Q	BD-Rate	TS_Q
文献[14]	HM 16.15	0	14.4%	0.04%	15.8%	0.12%	12.9%
文献[18]	HM 11.0	0.09%	14.4%	0.11%	15.8%	0.08%	14.9%
文献[25]	HM 16.0	—	—	2.51%	—	—	—
文献[26]	HPM 4.0	—	—	0.3%	27.6%	0.4%	30.6%
本文方法	HPM 4.0	0.23%	31.53%	0.25%	28.46%	0.27%	28.56%

[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]	Xiang MENG, Haibing YIN, Xiaofeng HUANG. Statistical modeling based fast rate distortion estimation algorithm for HEVC [J]. Telecommunications Science, 2021, 37(1): 58-68.
[3]	Liping ZHAO,Tao LIN,Kailun ZHOU,Hua PENG,Keli HU. A byte-size multi-variable-length-code based string matching algorithm for Alpha image coding [J]. Telecommunications Science, 2018, 34(11): 96-104.
[4]	Jinlong LI,Fen CHEN,Zongju PENG,Gangyi JIANG,Mei YU. Rate distortion optimization based on saliency characteristic for HEVC intra coding [J]. Telecommunications Science, 2017, 33(10): 124-133.
[5]	Siwei MA,Falei LUO,Tiejun HUANG. Kernel technologies and applications of AVS2 video coding standard [J]. Telecommunications Science, 2017, 33(8): 1-15.

SIMD-based parallelized rate distortion optimized quantization for AVS3

RichHTML

PDF下载

Knowledge

Abstract

Cite this article

share this article

Figures/Tables 16

References 29

Related Articles 5

Metrics

Recommended 0