Optimal selection of encoding configuration for scalable and multiple descriptionvideo coding

It is well-known that the wider the range of extraction points a scalable bitstream supports, the lower the compression efficiency at these extraction points. Moreover, this compression efficiency generally varies according to what combination of scalability types are used to support this range of extraction points as specified by the encoding configuration. Hence, we propose some objective criteria as a measure of coverage, compression efficiency and rate-distortion performance of a configuration, and then present a multiple-objective optimization formulation to select the best encoding configuration for scalable video coding, given a range of bitstreams that must be supported.Additionally, we present methods for scalable multiple description coding (SMDC) of monocular and multi-view video, where each description is scalable so that they can separately be streamed over DCCP with efficient rate adaptation. These descriptions are derived from an SVC-compliant bitstream, where there are variables related to the SVC encoding configuration and MD generation method used. We present a multiple objective optimization (MOO) framework to determine these variables so that the resulting SMDC strikes the best balance between maximizing the range of extraction points (coverage) of individual scalable descriptions, maximize the average end-to-end rate-distortion performance over the range of extraction points for a set of packet loss probabilities, and minimize the redundancy among descriptions. We performed optimization over the base layer rate (quantization parameter) and a selection of MD generation methods that feature various levels of redundancy at a fixed total rate for all descriptions. The framework is generic to allow optimization over other encoding variables as well if desired. Results of Monte-Carlo simulation of SMDC streaming of both monocular and stereoscopic videos demonstrate the performance of the proposed optimization framework.