Instructor Gaze Extraction for Accurate Work Attention Mapping
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing attention extraction systems fail to accurately capture and provide appropriate attention information based on gaze time, gaze transition, and attentive actions of instructors, making it difficult to effectively transfer skills and knowledge.
Innovation Solution
An attention extraction system that acquires video and coordinate information to identify gaze modes, sets gazed regions, and extracts images based on these modes, associating them with database information to determine right/wrong work targets and provide response information.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional line-of-sight tracking is used, then the system can track gaze position, but it cannot accurately extract attention information based on gaze time and gaze transition
Solution Approach 1:
The system performs preliminary actions by recording video information and coordinate information in time series order before extracting attention information. This allows the system to capture gaze time and gaze transition data that would otherwise be lost, enabling accurate extraction of attention information based on temporal patterns and sequences of gaze movements.
Solution Approach 2:
The system segments the gaze data into discrete time points and transitions, analyzing each segment separately to understand attention patterns. By dividing the continuous gaze stream into measurable segments with specific time intervals and coordinate changes, the system can accurately measure gaze time and transition characteristics to extract meaningful attention information.
2Measurement precision
If simple gaze position tracking is used, then the system structure is simple, but it cannot identify attentive actions and gaze modes
Solution Approach 1:
The system introduces dynamics by analyzing changes in gaze position over time rather than static positions. It identifies gaze modes (such as fixation, saccade, and smooth pursuit) and attentive actions by detecting dynamic patterns in the time series coordinate data, enabling the system to understand not just where the gaze is but how it moves and what that movement signifies.
Solution Approach 2:
The system adds temporal dimension to the gaze tracking by recording coordinate information in time series order. This transforms the data from two-dimensional spatial coordinates to three-dimensional spacetime data, enabling the system to analyze gaze transitions and identify attentive actions that would be invisible in static position tracking alone.
3Measurement precision
If video information is recorded continuously, then gaze transition information is captured, but data processing time increases
Solution Approach 1:
The system extracts only the necessary information from the continuous video and coordinate data by identifying specific gaze modes and attentive actions. Rather than processing every frame uniformly, it selectively extracts data points that correspond to meaningful gaze transitions and attention events, reducing processing time while maintaining accuracy in capturing gaze transition information.
Solution Approach 2:
The system processes data in periodic time intervals rather than continuously, analyzing gaze patterns at specific sampling rates. This periodic processing approach captures gaze transition information effectively while reducing the total processing time compared to continuous real-time analysis, balancing data completeness with processing efficiency.
Data Source
AI summary
In an attention extraction system (100) for extracting attention information on work, an attention extraction device (1) includes an acquisition device that acquires video information and coordinate information in time series order in association with the work, an identification device that obtains a viewpoint displacement of an instructor in a visual field range to identify a gaze mode of the instructor, an extraction device that sets a gazed region based on the gaze mode, and extracts a gazed image of a work target that the instructor has gazed at in the gazed region, and a storing device that associates the extracted gazed image with the visual field range, the viewpoint displacement, and the gaze mode, and stores them in a database as attention information on the work.