OptVV: An Adaptive Optimization Framework for Volumetric Video Streaming

The growing demand for interactive and personalized visual experiences, particularly in the metaverse era, highlights the importance of three-dimensional volumetric video. However, its large size and complex processing requirements pose significant challenges for practical streaming, often leading to resource waste and quality issues. To address this, we propose OptVV, an optimization framework for high-quality and bandwidth-efficient volumetric video streaming. The framework comprises three key components: (1) Adaptive QoE optimization, dynamically adjusting video quality based on network conditions, playback requirements, and volumetric video characteristics to ensure smooth playback; (2) an optimized DASH scheme for enhanced quality adaptation and bandwidth efficiency; and (3) optimal resource scheduling for real-time decoding, efficiently managing computational resources to maintain playback smoothness. Evaluation results show that OptVV significantly improves streaming performance, achieving up to 80% bandwidth savings, 177% higher viewport quality, 83% fewer playback stalls, and 72% faster decoding time compared to existing methods.