Low-Complexity Scalable Video Coding through Table Lookup VQ and Index Coding

摘要: The Internet community is very heterogeneous in terms of access bandwidth and terminal capabilities, hence, there much interest for low-computation, software-only, scalable video coders that guarantee universal to communication. Scalability allows users achieve a fair quality service relation their resources. Low complexity, on the other hand, necessary order ensure also with low computing power can be served.In this work, we propose multiplication-free codec, whose complexity reduced respect standard at price limited increase memory requirements. To end resort simple coding tools such as table lookup vector quantization (VQ) conditional replenishment. We start from coder proposed [1], which already guarantees high scalability computational burden, improve upon it by further reducing well encoding rate, no effect quality. main innovation use ordered VQ codebooks, encoder generate correlated indexes, unlike conventional VQ. Index correlation, turn, us carry out replenishment (the most time-consuming operation original coder) working indexes rather than block pixels, reduce drastically its complexity. In addition, take advantage correlation among compress them means predictive scheme, leads 15-20% rate reduction base layer, without significant Thanks these minor optimizations have obtained improved performance and, more important, 60-70% time (on general purpose machine) [1].