Exercise Feedback Display Using Gaze-Based Posture Comparison
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing home training systems face challenges in accurately analyzing user posture due to excessive recorded data and difficulty in maintaining focus on coaching images, leading to inaccuracies in posture imitation.
Innovation Solution
An electronic apparatus that considers a user's gaze direction to generate feedback images by comparing coaching and user images, using skeleton data and posture information, and stores these images for analysis when the gaze is not directed towards the device.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If all user exercise images are recorded for posture analysis, then the user can correct wrong posture using recorded images, but the amount of recorded data increases and all images must be viewed from the beginning unnecessarily
Solution Approach 1:
The system extracts only the essential information needed for posture analysis by comparing user images with coaching images in real-time, rather than recording all raw images. This extraction approach captures only the relevant posture data needed for correction, eliminating unnecessary data storage.
Solution Approach 2:
The system performs preliminary posture comparison and analysis during the exercise session itself, identifying posture issues as they occur. This preliminary action allows the system to flag only the problematic moments for review, rather than requiring review of the entire exercise session.
2Loss of information
If the user watches coaching images to learn proper posture, then the user can understand correct exercise form, but the user's eyes may not be directed to the electronic apparatus while watching coaching images, making it difficult to accurately imitate posture
Solution Approach 1:
The system provides real-time feedback by comparing the user's posture with the coaching image and highlighting discrepancies. This immediate feedback loop allows users to self-correct posture issues without needing to maintain constant visual focus on the electronic apparatus, as the system actively monitors and guides them.
Solution Approach 2:
The system enables users to perform self-assessment and self-correction of posture by providing automated comparison and feedback mechanisms. Users can independently identify and correct their posture deviations without requiring constant instructor intervention or intense concentration on the display device.
3Productivity
If the user makes motion while looking ahead to watch expert motion made while looking at side in coaching image, then the user can follow the coaching video, but the user would feel that their motion is different from the expert's motion
Solution Approach 1:
The system dynamically adjusts the comparison between user motion and expert motion based on real-time posture analysis. Rather than requiring exact visual replication, the system adapts the feedback to account for individual differences in motion execution, allowing users to maintain natural head and eye positions while still achieving accurate posture alignment.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
An electronic apparatus includes: a camera; a display; a memory; and a processor configured to: control the display to display a coaching image received from an external server and a user image obtained through the camera; identify a gaze direction of a user based on the obtained user image; based on the gaze direction of the user not being in a threshold range, obtain a feedback image including the coaching image and the user image based on posture information of a first object included in the coaching image and posture information of a second object included in the user image, and store the obtained feedback image in the memory; and control the display to display the stored feedback image.