A System Architecture for Live Immersive 3D-Media Transcoding over 5G Networks


Tele-Immersive (TI) applications are network based immersive media applications that enable multi-party real-time interactions of users located in different parts of the globe by placing them inside a shared virtual world. For each user participating in a TI application, their appearance is captured via multiple RGB-D sensors and a “3D replica” is created by utilizing full-body 3D-reconstruction algorithms that produce the person’s digitized 3D time varying mesh (TVM). TI applications produce large volumes of visual data in the form of multi-textured geometry, thus, creating a challenging networking scenario. Although TVM geometric data can be compressed per-frame via state-of-the-art static mesh compression algorithms, these techniques lack optimal performance since they do not exploit correlations of the data over time. As of today, TVM compression algorithms leveraging temporal exploitation of TVMs are very few and these TVM compression schemes are not quite ready yet to support real-time applications. Apart from the geometry aspect and in the simplest scenario, the texture components of TVMs are typically encoded as static images via typical image compression algorithms like JPEG2000. Thus, in order to cope with the large volume of data produced by a TI application, the invention of novel, efficient, real-time, and potentially adaptive, TVM compression schemes is highly demanded. However, due to the difficulties of developing such decent TVM codecs conforming to the aforementioned characteristics, an upgrade to a more efficient network infrastructure appears to be an additional necessity. TI applications have very high bandwidth requirements and ultra-low latency to allow for interactions and VR viewing as well as, in the future, delivery in high crowd-density areas. These comprise some of the main targets to be addressed by modern 5G networks.

  • A. Doumanoglou, N. Zioulis, D. Griffin, J. Serrano, T. Khoa Phan, D. Jimenez, D. Zarpalas, F. Alvarez, M. Rio, P. Daras, "A System Architecture for Live Immersive 3D-Media Transcoding over 5G Networks", Workshop on Media delivery innovations using flexible network models in 5G, IEEE International Symposium on Broadband Multimedia Systems and Broadcasting (BMSB18), Valencia, Spain, 6th – 8th June 2018

  • Visual Computing Lab

    The focus of the Visual Computing Laboratory is to develop new algorithms and architectures for applications in the areas of 3D processing, image/video processing, computer vision, pattern recognition, bioinformatics and medical imaging.

    Contact Information

    Dr. Petros Daras, Principal Researcher Grade Α
    1st km Thermi – Panorama, 57001, Thessaloniki, Greece
    P.O.Box: 60361
    Tel.: +30 2310 464160 (ext. 156)
    Fax: +30 2310 464164
    Email: daras@iti.gr