Scalable MDC-Based WebRTC Streaming for One-to-Many Volumetric Video Conferencing

Abstract

Video consumption has become central to modern life, with users seeking more immersive experiences such as virtual conferencing or concerts within virtual reality (VR). While 360° video offers rotational movement, it lacks true positional freedom. Fully immersive formats like light fields and volumetric video enable six degrees-of-freedom (6DoF), allowing both types of freedom. However, their high bandwidth and computational demands make them impractical for low-latency applications. Efforts to address these issues through compression and quality adaptation have improved quality of experience (QoE), but real-time interaction remains limited because of latency. To solve this, we introduce a novel, open-source one-tomany streaming architecture using point cloud-based volumetric video. By compressing point clouds with the Draco codec and transmitting via web real-time communication (WebRTC), we achieve low-latency 6DoF streaming. Content is adapted by employing a multiple description coding (MDC) strategy which combines sampled point cloud descriptions using the estimated bandwidth returned by the Google congestion control (GCC) algorithm. MDC encoding scales more easily to a larger number of users compared to individual encoding. Our proposed solution achieves similar real-time latency for both three and eight clients (163 ms and 166 ms), which is 9% and 19% lower compared to individual encoding. The MDC-based approach, using three workers, achieves similar visual quality compared to a per client encoding solution using five worker threads, and increased quality when the number of clients is greater than 20. Additionally, when compared to an approach with five fixed quality levels, our MDC-based approach scores 13% better in terms of latency, while achieving similar quality.