Toward optimality in scalable predictive coding

Abstract

Abstract-A method is proposed for efficient scalability in predictive coding, which overcomes known fundamental shortcomings of the prediction loop at enhancement layers. The compression efficiency of an enhancement-layer is substantially improved by casting the design of its prediction module within an estimation-theoretic framework, and thereby exploiting all information available at that layer for the prediction of the signal, and encoding of the prediction error. While the most immediately important application is in video compression, the method is derived in a general setting and is applicable to any scalable predictive coder. Thus, the estimation-theoretic approach is first developed for basic DPCM compression and demonstrates the power of the technique in a simple setting that only involves straightforward prediction, scalar quantization, and entropy coding. Results for the scalable compression of first-order Gauss-Markov and Laplace-Markov signals illustrate the performance. A specific estimation algorithm is then developed for standard scalable DCT-based video coding. Simulation results show consistent and substantial performance gains due to optimal estimation at the enhancement-layers.

QoS Support for Multi-user Sessions in IP-based Next Generation Networks

“…The former is composed of three flows with the same QoS requirements. Although MUSC can handle any number of flows, three flows allow a good trade-off between quality and bandwidth, and additional flows only provide marginal improvements [39]. Additionally, each flow has different priorities and exponential rates, which are common in scalable CODECs [39].…”

“…Although MUSC can handle any number of flows, three flows allow a good trade-off between quality and bandwidth, and additional flows only provide marginal improvements [39]. Additionally, each flow has different priorities and exponential rates, which are common in scalable CODECs [39]. Each one of the three flows has a Constant Bit Rate (CBR) of 32, 64 and 128 kb/s, starting from the most important to the less important one, respectively.…”

Transform-domain temporal prediction in video coding: Exploiting correlation variation across coefficients

“…The reason for the direct subtraction of the reference pixel block from the current block is that the temporal correlation coefficient, as appears between pixel blocks along the same motion trajectory, is typically close to 1. An alternative viewpoint that transform coefficients of the blocks in a motion trajectory form a scalar AR process, at each spatial frequency, was inspired by [4] where we proposed an estimation-theoretic (ET) approach to delayed video decoding, and [10] where an ET approach for predictive scalable coding was proposed. In both [4] and [10], such a viewpoint was necessary to explicitly account for the quantization interval information exploited by the ET framework, which is available only in the transform domain.…”

“…An alternative viewpoint that transform coefficients of the blocks in a motion trajectory form a scalar AR process, at each spatial frequency, was inspired by [4] where we proposed an estimation-theoretic (ET) approach to delayed video decoding, and [10] where an ET approach for predictive scalable coding was proposed. In both [4] and [10], such a viewpoint was necessary to explicitly account for the quantization interval information exploited by the ET framework, which is available only in the transform domain. The innovation of each scalar AR process (per frequency) was modeled as Laplacian [1], and the temporal correlation coefficient of each AR process at different spatial frequencies was assumed to be unity (as is the common practice in pixel domain).…”

Scalable predictive coding by nested quantization with layered side information

“…The enhancement layer is represented by coding residual error with respect to the current base-layer reconstruction, which results in very low coding efficiency. Rose and Regunathan [1] proposed a multiple-MCP-loop approach, in which the enhancement-layer predictor is optimally estimated by considering all the available information from both base and enhancement layers. However, closed-loop prediction has the disad- vantage of requiring the encoder to generate all possible decoded versions for each frame, so that each of them can be used to generate a predictor residue.…”