Project Details
Overview
 
 
 
Project description 

The rapid growth in terms of communication bandwidth, resolution, display modalities and mobility is calling for new types of multimedia communication systems. One of these new solutions, called Distributed Video Coding (DVC), has attracted attention from the research community during the past decade. Opposite to conventional video compression systems that feature complex encoding but simple decoding, DVC offers a reversed distribution with simple encoding but complex decoding. Such a scheme has an important advantage in up-link scenarios where encoders are complexity constrained. However, despite current efforts there is still a gap in compression performance between conventional solutions, such as H.264/AVC and DVC systems in the literature.
In this project we will improve DVC compression performance by focusing on various aspects. First of all, the accuracy of the models describing the relation between the original data at the encoder and the prediction generated at the decoder will be improved. Secondly, new architectures will be designed that incorporate a richer and more powerful set of coding modes, while preserving low-complexity at the encoder. Finally, research will be conducted towards algorithms that optimize quality in the context of constraints on rate, complexity, and delay.

Runtime: 2012 - 2015