Binary Arithmetic Coding System with Adaptive Probability Estimation by "Virtual Sliding Window"

Abstract

Virtual Sliding Window" algorithm presented in this paper is an adaptive mechanism for estimating the probability of ones at the output of binary non-stationary sources. It is based on "Imaginary sliding window" idea proposed by B.Ryabko. The proposed algorithm was used as an alternative adaptation mechanism in Context-Based Adaptive Binary Arithmetic Coding (CABAC) -an entropy coding scheme of H.264/AVC standard for video compression. The "virtual sliding window" algorithm was integrated into an open-source codec supporting H.264/AVC standard. Comparison of the "virtual sliding window" algorithm with the original adaptation mechanism from CABAC is presented. Test results for standard video sequences are included. These results indicate that using the proposed algorithm improves rate-distortion performance compared to the original CABAC adaptation mechanism. Besides improvement in rate-distortion performances the "Virtual Sliding Window" algorithm has one more advantage. CABAC uses a finite state machine (FSM) for estimation of the probability of ones at the output of a binary source. Transitions for FSM are defined by a table stored in memory. The disadvantage of CABAC consists in frequent reference to this table (one time for every binary symbol encoding), which is critical for DSP implementation. The "Virtual Sliding Window" algorithm allows to avoid using the table of transitions 1 .

Entropy Coding in HEVC

“…Note that this estimation is equivalent to using a sliding window technique [65,9] with window size W α = (1 − α) −1 . In the design of CABAC, Eq.…”

“…For every non-skipped CU, a regular coded flag called pred mode flag is sent to indicate the prediction mode, i.e., the decision whether the CU is either intra coded or inter coded. 9 Every non-skipped CU may be further subdivided into PUs, as shown for the example in Fig. 6.…”

An efficient multiplication-free and look-up table-free adaptive binary arithmetic coder

“…This task can be solved by utilizing several state machines with different adaptation speeds and precision of probability estimation [4], which unfortunately leads to an increase in memory consumption for storing different look-up tables. Another solution can be based on look-up table-free approach based on virtual sliding window [8] (VSW). In this approach, the probability estimation is calculated using a simple rule with one parameter -window length, and a trade-off is reached by assigning a specific window length selected according to the statistical properties of the corresponding binary source.…”

“…This random variable should take the same values at the corresponding steps of the encoder and the decoder. However, there is a way to avoid generating a random variable [8]. At step 1 of the algorithm let us replace a random value y t with its probabilistic average.…”

“…Based on (9), a probability estimation using virtual sliding window was proposed in [8]. In comparison with the Mcoder, it provides a better compression efficiency due to an assignment of a specific virtual sliding window length selected according to the statistical properties of the corresponding binary source.…”