Video coding for IP transmission by using polyphase downsampling multiple description

Abstract

When transmitting data over heterogeneous networks, there may be situations which necessitate dropping packets, for example, to accommodate a lower capacity or when congestion occurs. If retransmission of the lost packets is not feasible (e.g. for real time applications), we have to design a decoder able to recover significant information from the received packets. The literature suggests resolving this issue by using generalized multiple description coding (GMDC): it is source coding for multiple channels such that a decoder, which receives an arbitrary subset of the channels, may produce a useful reconstruction. In this work, we present a multiple description (MD) system based on a polyphase downsampling algorithm that splits an image source into an arbitrary number of balanced descriptions. Our goal is to obtain both the best reconstruction quality from the received descriptions (also only one) and a subjective indistinguishable reconstruction from the original if all the descriptions are available. Our specific interest is the communication of still images, but we think that our method is advantageous in other application such as video and audio.