Environment-Reactive Avatar Rendering for Realistic Mobile VR
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing avatar generation systems fail to accurately represent users, require high-performance hardware, and do not work well on power-constrained devices, often appearing cartoonish and not reflective of reality.
Innovation Solution
A technique for generating and rendering realistic avatars by embedding environmental properties like wind, rain, and lighting into the process, adjusting persona geometry based on environmental features, and combining tracking data with environmental conditions to create a photorealistic virtual representation that reacts to the scene.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If existing avatar generation systems are used, then avatars can be generated, but they appear cartoonish and do not accurately represent the user
Solution Approach 1:
The patent uses photogrammetry to capture real-world photographic images of the user and creates a digital replica or copy of their physical appearance. This copying approach directly addresses the cartoonish appearance problem by reproducing actual photographic data rather than generating synthetic 3D models, thereby improving both measurement precision of user features and manufacturing precision of realistic appearance.
Solution Approach 2:
The patent replaces traditional manual 3D scanning and modeling mechanical processes with automated photogrammetry software processing. By substituting the mechanical scanning system with a software-based image processing system that automatically generates 3D models from photographs, the system achieves higher accuracy in capturing user features while reducing the complexity of the generation process.
2Manufacturing precision
If high-performance general and graphics processors are used for avatar generation, then better avatar quality is achieved, but power consumption increases making it unsuitable for mobile devices
Solution Approach 1:
The patent performs the computationally intensive photogrammetry processing and 3D model generation as a preliminary action during an initial setup phase, rather than continuously during runtime on mobile devices. By completing the heavy computational work beforehand when high-performance resources are available, the system creates a finalized avatar model that can then be efficiently rendered on power-constrained mobile devices without requiring sustained high power consumption.
Solution Approach 2:
The patent uses readily available consumer camera photographs as the input data source instead of requiring specialized expensive scanning equipment. By utilizing ordinary photographic images that can be captured with mobile device cameras, the system achieves accurate avatar generation without needing continuous access to high-performance hardware resources, thereby reducing power consumption requirements for runtime operation.
Data Source
AI summary
Generating an avatar representation of a user involves obtaining scene data, determining environmental features, and obtaining tracking data for the user. An environmentally-adjusted geometric representation of the user is generated and presented in the scene. The environmentally-adjusted geometric representation is generated based on the tracking data and the environmental features, and is used to reflect environmental features of the scene in the virtual representation of the user. The environmentally-adjusted geometry representation enables a view of the avatar that appears to reflect physical characteristics of the scene in which the avatar is presented.