Gait Data Generation Device for Rehabilitation Learning Models
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for generating learning models using gait data require extensive rehabilitation data collection, especially when the response variable is a continuous value, which is time-consuming and labor-intensive.
Innovation Solution
A data generation device that acquires pair data combining gait data and response variables, generates measurement dataset vectors, and creates pseudo dataset vectors using covariance matrices, enabling the extension of datasets even when the response variable is continuous.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If rehabilitation data is collected to generate a learning model with sufficient accuracy, then the accuracy of the learning model is improved, but the time and effort required for data collection increases enormously
Solution Approach 1:
The patent performs preliminary data processing by extracting feature amounts from measurement gait data and generating covariance matrices in advance. This preliminary action creates a foundation that enables subsequent pseudo data generation without requiring extensive new data collection, thereby reducing the time and effort needed while maintaining model accuracy
Solution Approach 2:
The patent generates pseudo gait data by copying and transforming existing measurement gait data through mathematical operations (adding noise, applying covariance matrices). These synthetic copies serve as additional training data, eliminating the need to collect large amounts of real rehabilitation data while preserving the statistical properties needed for accurate learning models
2Adaptability or versatility
If the response variable is a continuous value, then the applicability of the data extension method is improved, but the ability to associate the response variable with extended explanatory variables deteriorates
Solution Approach 1:
The patent transforms the response variable from its original continuous form into a probabilistic distribution characterized by mean and covariance matrices. By changing the parameter representation and using statistical transformations (Cholesky decomposition, random sampling from multivariate normal distribution), the patent maintains the continuous nature of the response variable while enabling reliable association with pseudo-generated explanatory variables
Data Source
AI summary
Provided is a data generation device that acquires pair data constituted by a combination of measurement gait data relating to sensor data measured in accordance with the movement of a user's feet and a response variable corresponding to the measurement gait data, generates a measurement dataset vector by combining the response variable and a feature amount vector calculated using a feature amount extracted from measurement gait data, generates a covariance matrix relating to a plurality of pair data, generates pseudo gait data using the measurement gait data, generate a pseudo dataset vector by combining a pseudo feature amount vector calculated using a pseudo feature amount extracted from the pseudo gait data, and a pseudo response variable generated using a covariance matrix relating to the pseudo feature amount vector, and outputs the dataset.