Read the interview with Prof. Dr. Amr Rizk and his winning proposal: Exploring Real-time Transcoding of V-PCC Streams as a Service (RABBIT@SCALE)

Prof. Dr. Amr Rizk and Leibniz University Hannover
Prof. Dr. Amr Rizk is part of the Institute for Communication Technology at the Leibniz University Hannover which focuses on the design, performance evaluation and optimisation of networked systems. One of the main research areas of the Institute is multimedia communications with a focus on end-to-end immersive telepresence systems. Amr Rizk is the recipient of a number of Best Paper Awards in the area of multimedia systems. Recent research by his group includes methods for real-time transcoding and neural compression of point cloud data.
Can you give a brief overview of your winning proposal?
What are its key objectives and innovative aspects?
The project RABBIT@SCALE aims to develop a transcoding service for volumetric dynamic point-cloud streams that are encoded by MPEG-standardised Video-based Point Cloud Compression (V-PCC) and are consumed by adaptive bitrate streaming clients. The transcoding service can be leveraged within an immersive telepresence system to deliver volumetric streams to users with time-varying bitrates over 5G. Given the SPIRIT infrastructure, the project aims to evaluate the scalability of the transcoding service in terms of the number of clients while using multiple configurations and different software and hardware-based techniques. The project will be demonstrated in a real-world environment showcasing scalability for immersive telepresence systems.
What motivated you to apply for the SPIRIT Open Call?
The SPIRIT project offers through its Open Call a unique opportunity to bring the extensive research and development effort behind the transcoding service RABBIT into a real-world deployment at large scale. The SPIRIT project platform is ideal to experiment and test the boundaries of transcoding scalability for immersive telepresence given heterogeneous user conditions. The RABBIT@SCALE project aligns perfectly with the objectives of the SPIRIT project and the open call in terms of validating this third-party immersive streaming application on the SPIRIT infrastructure, as well as, providing new requirements for future developments through feedback to gauge the required SPIRIT resources.
How do you envision this project making an impact?
We envision that the project RABBIT@SCALE makes a significant impact demonstrating technological abilities of collaborative telepresence at large scale. From a research perspective, the experiment provides a demonstration of high-performance real-time transcoding of volumetric content rooted in the video-based encoding. This experiment establishes a use case that extends the currently available telepresence applications. While the transcoding-as-a-service pipeline is geared towards video-based encoding, it can serve as a template for showcasing transcoding-as-a-service with multiple core-transcoders that replace or work in parallel to the RABBIT transcoder. The experiment intends to evaluate the trade-offs between quality and transcoding speed while scaling the number of clients to show the ability of SPIRIT to support collaborative telepresence at large scale. Researchers using the SPIRIT platform may utilise the developed framework beyond this project to test combinations of different transcoding services and adaptive bitrate streaming algorithms.