Lossless VR Scene Transmission via Pre-rendered Perspective Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current virtual reality (VR) streaming technologies face limitations in providing high-quality, immersive experiences due to data compression issues, which result in pixilation, jitter, and restricted perspectives, limiting user interaction and immersion across distributed devices.
Innovation Solution
Transmitting 3D scene data and tracking information from a source device to client devices before rendering, allowing for lossless transmission and flexible perspective adjustment using virtual cameras, enabling multiple perspectives and improved interaction within shared VR environments.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of energy
If VR video data is compressed to preserve network bandwidth during streaming, then bandwidth usage is reduced, but image quality degrades resulting in pixilation, jitter, and visual artifacts
Solution Approach 1:
The system performs pre-rendering of VR scenes from multiple perspectives before streaming. By preparing multiple perspective views in advance at the source device, the system can transmit these pre-rendered perspectives to recipient devices without requiring real-time compression, thereby maintaining high image quality while managing bandwidth through selective transmission of only necessary perspective data.
2Adaptability or versatility
If VR video is rendered on a source device before streaming to recipient devices, then the VR experience can be shared across multiple users, but the receiving devices have limited ability to change the perspective or field of view since the perspective is baked in to the rendered video
Solution Approach 1:
The system segments the VR scene rendering by generating multiple discrete perspective views from a single 3D environment. Instead of rendering one fixed perspective, the source device creates several distinct perspective representations that can be independently selected and transmitted to different recipient devices, allowing each user to receive the perspective data most suitable for their viewing needs.
Solution Approach 2:
The pre-rendered multiple perspectives serve multiple functions: they can be transmitted to multiple recipient devices simultaneously, each device can select the most appropriate perspective for its user, and the system maintains adaptability for different viewing scenarios without requiring complex real-time rendering at recipient devices.
3Manufacturing precision
If large quantities of VR data are streamed to recipient devices to preserve high refresh rate and resolution, then image quality is maintained, but network bandwidth requirements increase significantly
Solution Approach 1:
By pre-rendering multiple perspectives at the source device before streaming, the system prepares high-resolution data in advance. This allows recipient devices to receive pre-processed perspective data that maintains high resolution and refresh rate without requiring the source device to continuously stream massive amounts of raw VR data, thereby reducing real-time bandwidth requirements while preserving image quality.
Data Source
AI summary
Systems, methods, and computer-readable media herein provide multiple perspectives for viewing lossless transmissions of VR scenes. Tracking and scene data associated with a three dimensional “3D” scene are received and used to determine viewing parameters that define a particular perspective from which to render the 3D for display at a particular recipient device. The tracking and scene data is streamed without compression to recipient devices prior to rendering allowing for dynamic and flexible viewing perspective selection by the recipient devices. The recipient devices may render and display the 3D scene in accordance to a particular viewing perspective.