Markerless Motion Capture with Distance-Adaptive Avatar Tracking
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing augmented reality (AR) software struggles with accurate and simultaneous tracking of body, face, and hand movements, especially when transitioning between close-range and far-range shots, and lacks the ability to efficiently capture and animate digital avatars in real-time across Web2 and Web3 environments.
Innovation Solution
AR software utilizing artificial intelligence to prioritize anchor points based on distance and employing a hybrid rendering approach that combines real-time local and delayed remote processing to enhance avatar animation and ownership verification through non-fungible tokens.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If AR software captures markerless motion data at close-range distances, then hand and face tracking precision is improved, but body tracking completeness deteriorates
Solution Approach 1:
The system dynamically adapts the set of tracked anchor points based on the detected distance between the camera and the user. At close-range distances, the system prioritizes tracking hand and face anchor points with higher precision. At far-range distances, the system expands tracking to include full body anchor points. This dynamic adjustment resolves the contradiction by allowing the system to optimize for either precision or completeness depending on the operational context.
Solution Approach 2:
The system applies different tracking qualities to different body regions based on distance. At close-range, high-precision tracking is applied locally to hands and face, while body tracking is reduced or omitted. At far-range, the tracking quality is distributed more evenly across the entire body. This local quality differentiation allows the system to maintain high precision where needed while adapting to distance constraints.
2Adaptability or versatility
If AR software captures markerless motion data at far-range distances, then body tracking completeness is improved, but hand and face tracking precision deteriorates
Solution Approach 1:
The system dynamically switches between close-range and far-range tracking modes based on detected distance. In far-range mode, the system expands the set of tracked anchor points to include the full body, accepting reduced precision for individual features. In close-range mode, the system focuses on high-precision tracking of hands and face. This dynamic mode switching resolves the contradiction by adapting tracking behavior to the operational distance.
Solution Approach 2:
The system employs partial tracking at far-range distances, focusing on capturing the essential body pose and movement rather than attempting to track every detail with high precision. This partial action approach allows the system to maintain body tracking completeness while accepting reduced precision for specific features like hands and face, which are harder to resolve at distance.
3Speed
If AR software processes motion data in real-time locally, then processing speed is improved, but computational resource consumption increases
Solution Approach 1:
The system segments the motion processing into distinct stages: real-time local processing of anchor point detection and basic motion capture, followed by delayed remote processing for high-quality rendering and complex computations. This segmentation allows the system to maintain real-time responsiveness for essential functions while offloading resource-intensive tasks to remote servers, thereby resolving the contradiction between processing speed and resource consumption.
Solution Approach 2:
The system introduces an intermediary hybrid rendering approach that combines local real-time processing with remote delayed processing. The local device handles time-critical motion capture and anchor point identification, while the remote server handles computationally intensive rendering and post-processing. This intermediary approach distributes computational load, maintaining speed for essential functions while managing overall resource consumption.
4Adaptability or versatility
If AR software tracks all body parts simultaneously, then motion capture completeness is improved, but system complexity increases
Solution Approach 1:
The system dynamically adjusts the number and type of anchor points tracked based on the detected distance. At close-range, the system tracks a limited set of high-priority anchor points (hands and face). At far-range, the system expands tracking to include a comprehensive set of body anchor points. This dynamic adjustment reduces system complexity at close-range while maintaining motion capture completeness at far-range, resolving the contradiction.
Solution Approach 2:
The system employs partial tracking at close-range distances, focusing on essential anchor points rather than attempting to track all body parts simultaneously. This partial action reduces the computational burden and system complexity while still capturing the most important motions. At far-range distances, the system transitions to more complete tracking when the additional computational resources become feasible.
Data Source
AI summary
A computing system captures markerless motion data of a user via a camera of the computing system. The computing system retargets the first plurality of points and the second plurality of points to a three-dimensional model of an avatar associated with the user, wherein the avatar is associated with an identity non-fungible token that uniquely represents the user across Web2 environments and Web3 environments, and wherein retargeting the first plurality of points and the second plurality of points animates the three-dimensional model of the avatar. The computing system renders a video local to the computing system, wherein the video comprises the markerless motion data of the user retargeted to the three-dimensional model of the avatar causing hands, face, and body of the avatar to be animated in real-time. The computing system causes a non-fungible token to be generated, the non-fungible token uniquely identifying ownership of the video.