Probability-Based Fast Intra Prediction Algorithm for Spatial SHVC

Abstract

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.

Gaussian Distribution-based Mode Selection for Intra Prediction of Spatial SHVC

“…In order to demonstrate the performance of the proposed algorithm, we compare the performance of our algorithm with FIICA [5] and PBFIP (%) [12]. According to common SHM test conditions (CSTC) [15], only one QP set is used for BL: (22, 26, 30, 34) and two QP sets for EL: (22, 26, 30, 34) and (24, 28, 32, 36).…”

“…Wang et al [11] make decisions on whether to skip the Intra mode or not by checking the distribution of the residual coefficients of ILR mode. Since the probabilities of modes are strongly relative to mode selection, Wang et al [12] combine the probability of ILR mode with its residual coefficients to early skip Intra mode. The works in [13] [14] adopt zero blocks to early terminate mode selection.…”

Hybrid Strategies for Efficient Intra Prediction in Spatial SHVC

“…Wang et al [8] first check ILR mode, and then decide whether its residual coefficients follow Gaussian distribution so as to early skip Intra mode. Wang et al [11] combine ILR mode probabilities with its residual coefficients to skip Intra mode prediction. Pan et al [12] combine depth correlation and all-zero block to early terminate mode selection.…”

“…Wang et al [8] combine relevant CUs with DMs and their corresponding HC values to predict candidate DMs and skip unlikely DMs. Wang et al [11] adopte the difference of Hadamard Costs (HCs) of typical DMs to predict candidate DMs, and then combine percentages of gradient amplitudes with HCs to early terminate DM selection. Jamali et al [17] develop an RDO cost statistical model to predict likely DMs.…”

“…As textural features are related to depth selection, using textual features to predict candidate depths and skip unlikely coding depths is a common practice. Wang et al [11] integrate depth probabilities with textural complexity to early skip unlikely depths and early terminated depth selection. The work in [18] employs a tunable decision model to predict candidate CU size partition in Versatile video coding.…”