The evolution of multiple description coding based on scalar quantization

Fabio Verdicchio, Adrian Munteanu, Jan Cornelis, Peter Schelkens

Research output: Contribution to conferencePaper

Abstract

Multiple Description Coding (MDC) allows for error resilient coding of data. This tutorial paper discusses one of the most popular MDC technologies, namely MDC based on scalar quantization, and its application in scalable coding of multimedia. This concept is illustrated by a fully scalable video coder, based on motion compensated temporal filtering, which incorporates the MDC principle for the encoding of both texture and motion vector data. For such system, we derive the principles of channel-aware MDC rateallocation, which enables adjusting both the bit-rate and the resilience to errors of the output stream. Such adaptation is performed on-the-fly depending on the channel condition. Experimental results prove the advantages of MDC and the effectiveness of MDC system when transmitting video streams over unreliable networks.
Original languageEnglish
Pages1990-1994
Number of pages5
Publication statusPublished - Sep 2007
EventEuropean Signal Processing Conference - Poznan, United Kingdom
Duration: 6 Sep 2007 → …

Conference

ConferenceEuropean Signal Processing Conference
CountryUnited Kingdom
CityPoznan
Period6/09/07 → …

Fingerprint Dive into the research topics of 'The evolution of multiple description coding based on scalar quantization'. Together they form a unique fingerprint.

  • Cite this

    Verdicchio, F., Munteanu, A., Cornelis, J., & Schelkens, P. (2007). The evolution of multiple description coding based on scalar quantization. 1990-1994. Paper presented at European Signal Processing Conference, Poznan, United Kingdom. http://www.eurasip.org/Proceedings/Eusipco/Eusipco2007/Papers/d2l-a01.pdf