Avatar Motion Blending With Fewer Sensors and Real-Time Pose Estimation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing motion capture techniques require multiple motion sensors attached to the body, which can be cumbersome and costly, and struggle to accurately estimate positions and orientations of non-attachment sites in real-time for effective blending of multiple actors' motions in virtual spaces.
Innovation Solution
A system using a reduced number of inertial sensors and combining inverse kinematics with inertial navigation and dynamics modeling to estimate positions and orientations of both attachment and non-attachment sites, employing a Kalman filter for correction to minimize errors over time.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If multiple motion sensors are attached to the body for motion capture, then measurement precision is improved, but device complexity and sensor load increase
Solution Approach 1:
The patent replaces the mechanical sensor attachment system with a hybrid computational approach combining inertial navigation system (INS) data with inverse kinematics (IK) calculations. Instead of using multiple physical sensors throughout the body, the system uses a minimal set of inertial sensors combined with mathematical models to estimate positions and orientations of all body segments, thereby substituting mechanical measurement with computational estimation.
Solution Approach 2:
The patent introduces inverse kinematics as an intermediary computational process that bridges the gap between limited sensor measurements and comprehensive motion capture. The IK algorithm acts as a mediator that takes INS data from a minimal sensor set and transforms it into full-body pose estimation, enabling accurate motion capture without requiring dense sensor placement.
2Reliability
If multiple motion sensors are attached to the body, then reliability of motion capture is improved, but ease of operation deteriorates due to cumbersome sensor attachment
Solution Approach 1:
The patent extracts and removes unnecessary motion sensors from the body while maintaining motion capture reliability through computational compensation. By using inverse kinematics to mathematically infer the positions of unmeasured body segments from limited sensor data, the system takes out redundant sensors that would otherwise be required for reliable measurement, thereby simplifying the attachment process.
Solution Approach 2:
The patent changes the fundamental approach from relying on multiple physical sensor parameters to using computational parameters through inverse kinematics. Instead of increasing sensor count to improve reliability, the system transforms the problem into a computational domain where mathematical models and algorithms provide the necessary reliability without additional physical components.
3Productivity
If existing motion capture techniques are used, then real-time reproduction is achieved, but accuracy of non-attachment sites deteriorates
Solution Approach 1:
The patent applies preliminary calibration and model setup actions during system initialization, where inverse kinematics models are pre-configured with subject-specific anatomical parameters and segment relationships. This preliminary preparation enables the system to perform accurate real-time estimation of non-attachment sites without requiring additional computational resources during actual motion capture, thus maintaining both speed and accuracy.
Data Source
Figure 1
Figure 2~3
Figure 4
AI summary
[Overview] [Problem to be Solved] To provide a program, an information processor, and an information processing method that make it possible to blend motions of a plurality of actors captured by using a motion capture technique and to reproduce the blended motions in real time in an avatar or the like on a virtual space. [Solution] There is provided a program that causes a computer to implement a control function of dynamically controlling a motion of an avatar in a virtual space or a robot on a real space, the control function being configured to: capture motions of a plurality of actors on the real space from respective motion sensors attached to the actors; blend the motions of the plurality of actors on the basis of a predetermined algorithm; and dynamically control the motion of the avatar or the robot on the basis of the blend result to cause the avatar or the robot to make a motion reflecting the motions of the plurality of actors.