Foot Motion Estimation From Upper-Body Trackers in VR
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current body tracking technologies require trackers to be attached to the feet, which is cumbersome for users.
Innovation Solution
A motion determination apparatus that specifies an operative foot and an inoperative foot based on sensing data from upper body trackers, determining foot motions with different coefficients for each foot, and using machine learning to estimate foot positions and directions without direct foot tracking.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If trackers are attached to the feet to precisely represent body motions, then measurement precision is improved, but ease of operation deteriorates due to cumbersome attachment
Solution Approach 1:
The invention extracts the foot tracking function from direct foot-mounted trackers and relocates it to upper body trackers. By measuring velocity of upper body regions (head, hands, torso) and using this data to infer foot motion through machine learning, the system eliminates the need for foot tracker attachment while maintaining motion representation precision
Solution Approach 2:
The invention introduces upper body velocity measurement as an intermediary between direct foot tracking and body motion representation. The velocity of upper body regions serves as a proxy that correlates with foot motion, allowing the system to determine operative vs. inoperative feet and calculate foot positions without direct foot contact
2Reliability
If trackers are attached to all body regions including feet, then reliability of body tracking is improved, but device complexity increases
Solution Approach 1:
The invention removes foot trackers from the required configuration, reducing the number of tracking devices from six (head, two hands, two feet, torso) to four (head, two hands, torso). The reliability is maintained through machine learning algorithms that infer foot motion from upper body velocity patterns
Solution Approach 2:
The system creates a virtual model of foot motion by copying and adapting upper body motion patterns. Through machine learning, the system learns the correlation between upper body velocity and foot motion, then uses this learned relationship to generate accurate foot position data without physical foot trackers
Data Source
AI summary
A motion determination apparatus, a learning apparatus, a motion determination method, a learning method, and a program capable of precisely representing motions of the overall body of a user without attaching trackers to the feet of the user are provided. A speed specification section specifies, on the basis of sensing data output from a tracker attached to a region of the upper body of a user, a direction of a speed of the region. An operative foot specification section specifies, on the basis of the specified direction of the speed, one of the left foot and the right foot of the user as an operative foot and specifies the other as an inoperative foot. A motion determination section determines, for each of the left foot and the right foot, a motion of the foot on the basis of whether the foot is the operative foot or the inoperative foot.