Early Intra CU Size Decision for Versatile Video Coding Based on a Tunable Decision Model

Li, Yue; Yang, Gaobo; Song, Yun S.; Zhang, Hanling; Ding, Xiangling; Zhang, Dengyong

doi:10.1109/tbc.2021.3073556

Cited by 48 publications

(22 citation statements)

References 38 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Table III indicates the proposed method achieves 48.5% encoding time saving on average with 0.59% BDBR increase on average. Table III shows the average BDBR values of Yang et al [4], Fu et al [28], Chen et al [29] and Li et al [24] increased by 1.56%, 0.38%, 1.54% and 1.36%, respectively. Meanwhile, their time saving are 52.6%, 28.8%, 51.0% and 46.2% for all video sequences, respectively.…”

Section: B Experimental Results Of the Proposed Methodsmentioning

confidence: 98%

“…The Table II and Table III provides the results of the proposed algorithm compared with the latest fast methods. We compared the proposed overall scheme with the latest fast methods of H.266/VVC, including those of Chen et al [19], Zhang et al [20] and Park et al [23] by using VTM 4.0 and the other latest fast methods of H.266/VVC, including those of Yang et al [4], Fu et al [28], Chen et al [29] and Li et al [24] by using VTM 2.0. These methods were efficient, BDBR increase of only 0.56%.…”

Section: B Experimental Results Of the Proposed Methodsmentioning

confidence: 99%

“…Park et al [23] proposed a simple early decision technique that can effectively reduce TT complexity. Li et al [24] proposed a tunable decision model based on the prediction distortion to achieve different trade-off between encoding loss and complexity reduction.…”

Section: Related Workmentioning

confidence: 99%

See 2 more Smart Citations

High Efficiency Intra CU Partition and Mode Decision Method for VVC

Lin

Chen

et al. 2022

IEEE Access

View full text Add to dashboard Cite

Versatile video coding (VVC/H.266) is the newest video compression standard, which is developed by the Joint Video Experts Team. Compared with previous encoding schemes, VVC achieves higher compression efficiency by introducing a new partition structure and additional intra prediction modes but results in high computational complexity. To efficiently solve this problem of redundant processing in quad-tree with nested multi-type tree structures and intra mode prediction, we propose a texture analysis-based ternary tree (TT) and binary tree (BT) partition strategy, and a gradient-based intra mode decision method to accelerate TT and BT partition and intra mode prediction, separately. The texture complexity and prediction direction of coding unit (CU) is calculated by texture detection method. A texture analysis-based TT and BT partition strategy is established by using the regression method based on analyzing the texture complexity of the CU. Then, a texture analysis-based TT and BT partition strategy is applied to reduce the redundant partition for each CU. By using the prediction direction of CU, a gradientbased intra mode decision method is established for skipping the impossible modes for each CU. Experimental results revealed that the proposed method could save 49.49% in encoding time and increase the Bjontegaard delta bit rate (BDBR) by only 0.56%. It confirms that the proposed method achieved high efficiency and a good balance between the BDBR and time saving.INDEX TERMS H.266/VVC, fast CU partition, intra mode decision, gradient-based, regression

show abstract

Section: B Experimental Results Of the Proposed Methodsmentioning

confidence: 98%

Section: B Experimental Results Of the Proposed Methodsmentioning

confidence: 99%

See 1 more Smart Citation

High Efficiency Intra CU Partition and Mode Decision Method for VVC

Lin

Chen

et al. 2022

IEEE Access

View full text Add to dashboard Cite

show abstract

“…The research in [4] splits residual coefficients of a coding unit (CU) into top and bottom parts as well as left and right parts, then decides whether to further split this CU according to the difference of two parts of each division. The work in [5] first obtains the difference between original luminance pixels and predicted ones, and then develop two decision models to early skip of vertical binary-tree and horizontal and vertical ternary-tree partition for each CU. The work in [6] first develops an RDO cost statistical model, and use HCs to predict candidate DMs and skip unlikely DMs based on the model.…”

Section: Related Workmentioning

confidence: 99%

Probability-Based Fast Intra Prediction Algorithm for Spatial SHVC

Wang

Sun

Liu

et al. 2022

IEEE Trans. on Broadcast.

View full text Add to dashboard Cite

Due to multi-layer encoding and Inter-layer prediction, Scalable High Efficiency Video Coding (SHVC) has extremely high coding complexity. It is very crucial to improve its coding speed so as to promote widespread and cost-effective SHVC applications. In this paper, we propose a new probabilitybased fast Intra prediction algorithm for spatial SHVC. More specifically, first, we integrate depth probabilities with textural based all-zero blocks and all-nonzero blocks through Lagrange Interpolation Polynomial (LIP) to derive thresholds to early skip unlikely depths and early terminate depth selection. Second, to early skip Intra mode prediction, we combine mode probabilities with Jarque-Bera test through LIP to test the residual coefficients of Inter-layer reference (ILR) mode. Third, we adopt the difference of Hadamard Costs (HCs) of horizontal, vertical and diagonal directional modes to predict candidate directional modes (DMs), and then exploit both the percentages of gradient amplitudes and HCs for DM early termination. Experimental results demonstrate that the proposed algorithm can achieve a speed up gain of more than 61% with 0.03% decrease in Bjøntegaard Delta Bit Rate on average.

show abstract

“…To improve the flexible QTMT structure, [39] proposed a multi-stage CU partition by a CNN model trained with an adaptive loss function. Li et al [40] established two decision models by exploiting the hierarchal correlation of the luma prediction distortion for early skipping Binary Tree (BT) and Ternary Tree (TT) partition. To reduce encoding complexity, Wu et al [41] trained split classifier and splitting directional classifier based on SVM for different sizes of CUs.…”

Section: Low-complexity Vvc Optimizationmentioning

confidence: 99%