Gaze Heat Map Estimation for Driver Concentration Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for monitoring driver attention fail to accurately estimate concentration levels under varying load conditions, leading to inefficiencies in providing timely warnings or assistance.
Innovation Solution
An electronic device equipped with an image-capturing unit, line-of-sight detector, and controller that utilizes machine learning to estimate concentration levels by generating low- and high-concentration heat maps based on line-of-sight data, accounting for different load factors and their removal.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If cumulative visibility method is used to monitor driver attention, then the system complexity is reduced, but the measurement precision of concentration estimation deteriorates
Solution Approach 1:
The patent segments the concentration estimation problem into multiple load factors (e.g., auditory load, visual load, cognitive load) and creates separate heat map estimators for each condition. This segmentation allows the system to handle different distraction scenarios independently, improving measurement precision without requiring a single overly complex unified model.
Solution Approach 2:
The patent performs preliminary machine learning training to generate heat map estimators for various load conditions before actual monitoring begins. By pre-computing the relationships between line-of-sight patterns and concentration levels under different loads, the system achieves high measurement precision during operation without real-time complex calculations.
2Measurement precision
If machine learning with multiple load factors is implemented, then the measurement precision of concentration estimation is improved, but the device complexity increases
Solution Approach 1:
The patent applies local quality by creating specialized heat map estimators tailored to specific load conditions (auditory, visual, cognitive loads). Each estimator is optimized for its particular condition rather than using a single generic model, which improves precision for each scenario while keeping individual estimators relatively simple in structure.
Solution Approach 2:
The patent changes parameters by training separate models for different load factors and transitioning between them based on detected conditions. The system adjusts which heat map estimator is active based on the current load state, allowing high precision across varying conditions without maintaining all complex models simultaneously in operation.
3Adaptability or versatility
If heat map estimation for multiple load factors is performed, then the adaptability to different distraction conditions is improved, but the loss of time for processing increases
Solution Approach 1:
The patent performs all machine learning training and heat map generation in advance during a preprocessing phase. By pre-computing the heat maps for various load conditions and storing them, the system achieves high adaptability to different distraction scenarios during operation without real-time processing delays, as the estimators are already prepared and ready for immediate use.
Data Source
AI summary
An electronic device 10 includes an image-capturing unit 11, a line-of-sight detector 12, and a controller 14. The controller 14 functions as a low-concentration heat map group estimator 15 and a high-concentration heat map estimator 16. The low-concentration heat map group estimator 15 can estimate a low-concentration heat map group based on an image. The high-concentration heat map estimator 16 can estimate a high-concentration heat map based on the image. The controller 14 calculates the degree of concentration of a subject based on the low-concentration heat map group, the high-concentration heat map, and a line of sight of the subject. The low-concentration heat map group estimator 15 is constructed using learning data obtained by machine learning the relationship between learning images and lines of sight for each load factor. The high-concentration heat map estimator 16 is constructed using learning data obtained by machine learning the relationship between learning images and lines of sight when the load has been removed or the load has been reduced.