The intended audience of this paper is one intimately familiar with scalable video coding (SVC) using wavelet transforms. In other words, the technical depth of the paper is most appropriate for experts in the field of wavelet-based SVC; it is inaccessible to the general public. More specifically, the paper builds on a motion compensated embedded zero block context modeling algorithm for scalable coding (MC-EZBC). It applies embedded block coding by optimized truncation (EBCOT) instead of EZBC on wavelet coefficients for texture coding, and improves the rate control through a hybrid approach: adapting the byte budget to the importance of temporal frames (allocating more for low subbands) and then using post compression rate distortion (PCRD) to avoid dividing bitrate equally across frames.
The core novelty and contribution of the paper lies in this rate control algorithm, which is an improvement on the original EZBC rate control where entire subbands are coded together without any rate distortion optimization. The adaptive nature of the proposed rate control algorithm, which assigns more bits to more important parts of the spatiotemporal frames, increases the compression efficiency of the particular wavelet-based SVC codec being built on.
The choices the researchers made in designing their improvements on the scalable codec are well justified and explained, and adequate performance evaluation is provided relative to similar approaches. Since the standardized H.264 SVC amendment does not use wavelet-based compression, the scope of this publication is limited to academic studies and research. Even within wavelet-based SVC research, this contribution is relevant to those studying MC-EZBC. Despite this narrowness, the novel contribution and the improvement in performance warrant a researcher in this specific area to read this paper.