AR Avatar Gesture Redirection with Dynamic Multi-Objective IK
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for mapping user body pose to telepresence avatars in mixed/augmented reality (MR/AR) systems result in erroneous positional interactions due to differences in physical dimensions and localization errors, conflicting with the need for accurate end effector redirection and preservation of original body pose.
Innovation Solution
A multi-objective optimization (MO) framework with dynamic weight allocation is employed to minimize positional errors and preserve the original body pose, using a primary objective to correct end effector positioning and a secondary objective to maintain the original interaction intent, adaptable to different error tolerances.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If direct one-to-one mapping of user joint angles to avatar is used, then mapping simplicity is maintained, but positional interaction accuracy deteriorates due to differences in physical dimensions and localization errors
Solution Approach 1:
The patent introduces an inverse kinematics (IK) solver as an intermediary between the user's joint angles and the avatar's joint angles. The IK solver computes the appropriate avatar joint angles that will position the avatar's end effector at the goal target, rather than directly mapping user joint angles. This intermediary computation resolves the contradiction by maintaining mapping simplicity while improving positional accuracy through mathematical transformation.
Solution Approach 2:
The patent changes the parameter mapping approach from direct joint angle mapping to end effector position-based mapping. Instead of mapping joint angles directly, the system maps the user's end effector position to the avatar's end effector position, then uses IK to compute the corresponding joint angles. This parameter transformation resolves the contradiction by decoupling the mapping simplicity from positional accuracy.
2Measurement precision
If end effector redirection is applied to minimize positional error, then interaction accuracy improves, but original body pose is lost deteriorating non-verbal cue preservation
Solution Approach 1:
The patent implements a dynamic weighting mechanism that adjusts the importance of different objectives based on the current error magnitude. When the initial positional error is large, the system prioritizes end effector redirection to the goal target. When the error is small, the system transitions to prioritizing body pose preservation. This dynamic adjustment resolves the contradiction by adaptively balancing accuracy and pose preservation based on real-time error conditions.
Solution Approach 2:
The patent applies preliminary redirection only when necessary - specifically when the initial error exceeds a threshold. The system first calculates the error between the user's end effector and the goal target, then conditionally applies IK-based redirection only if the error is significant. This preliminary assessment resolves the contradiction by avoiding unnecessary pose alterations when the error is already within acceptable bounds.
3Device complexity
If static weighting of objectives is used in IK approach, then computational simplicity is maintained, but adaptability to varying error magnitudes deteriorates
Solution Approach 1:
The patent implements dynamic weighting where the relative importance of end effector positioning versus body pose preservation changes based on the magnitude of the initial error. The system calculates the initial error, then uses this value to dynamically adjust the weights in the IK objective function. This dynamic approach resolves the contradiction by making the system adaptable to varying error conditions without requiring complex manual tuning of static weights.
Data Source
AI summary
A system and method for mapping a user's gestural interactions (positional interactions and pointing interactions) from their environment to that of their avatar's environment. The system and method provide a multi-objective optimization (MO) process with the primary objective being to minimize the positional error between the end effector and goal target and the secondary objective being to preserve the original body pose while redirecting the end effector to the goal target. The primary and second objectives are addressed with a dynamic weight allocation approach that optimizes between the objectives of body joint redirection and preserving the body pose of the original interaction.