Publications-Detail

High Quality Video Conferencing: Region of Interest Encoding and Joint Video/Audio Analysis

Authors:
Bulla, C. ,  Feldmann, C. ,  Schäfer, M.Heese, F.Schlien, T. ,  Schink, M.
Journal:
International Journal on Advances in Telecommunications
Volume:
6
Page(s):
153 - 163
number:
3 & 4
Date:
Dec. 2013
ISSN:
1942-2601
Language:
English

Abstract

In this paper, we present a high quality video conferencing system, that has been developed in the collaborative project “Connected Visual Reality (CoVR) – High Quality Visual Communication in Heterogeneous Networks” and was designed to reduce bitrate while preserving a constant visual quality. We utilize the fact that the main focus in a typical video conference lies upon the participating persons to save bitrate in less interesting parts of the video and introduce a scene composition concept that is merely based on the detected regions of interest. The region of interest encoding and the scene composition will be supported by a joint video and audio analysis. On the video analysis side we use a Viola-Jones face detector to detect, and a MeanShift tracker to track the regions of interest. The audio analysis exploits the information from the video analysis about the detected participants by a beamforming algorithm and creates an activity index for each participant. To represent the detected region of interests for the encoder we use a quality map on the level of macro-blocks, which allows the encoder to choose its quantization parameter individually for each macro-block. Finally, the proposed scene composition omits the background and shows only the most active participants of the conference, thus visual quantization artifacts introduced by the encoder get irrelevant. Experiments on recorded conference sequences demonstrate bitrate savings up to 50% that can be achieved with the proposed system.

Download

BibTeX

Copyright © by IKS
bulla13.pdf
This material is presented to ensure timely dissemination of scholarly and technical work. Copyright and all rights therein are retained by authors or by other copyright holders. All persons copying this information are expected to adhere to the terms and constraints invoked by each author's copyright. In most cases, these works may not be reposted without the explicit permission of the copyright holder.