Driver Gaze Distribution Comparison for Attentiveness Estimation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current driver monitoring systems require significant processing resources to analyze large amounts of image data for estimating driver state based on eye gaze patterns, leading to reduced response times and inaccurate results due to explicit detection of scene parameters.
Innovation Solution
A system utilizing an outward-facing camera to capture a video stream of the environment and an inward-facing camera to capture a driver's face, with a neural controller generating an expected gaze distribution and an eye tracker controller extracting actual gaze directions, calculating a distance measure between the two distributions using divergence algorithms to determine driver attentiveness without explicit object detection.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If driver monitoring systems process large amounts of image data to detect scene parameters explicitly, then measurement precision of driver state is improved, but processing time increases and response time decreases
Solution Approach 1:
The patent extracts only the essential gaze direction information from the driver's eyes without processing the entire scene. The system focuses specifically on eye gaze patterns rather than analyzing all image data, thereby reducing processing time while maintaining accuracy in driver state estimation.
Solution Approach 2:
The system segments the driver monitoring task into specific eye gaze detection rather than comprehensive scene analysis. By dividing the problem into focused eye tracking components, the system processes only relevant data portions, improving response time while preserving measurement precision for driver attentiveness.
2Measurement precision
If driver monitoring systems process large amounts of image data for explicit detection of scene parameters, then measurement precision is improved, but processing resources increase
Solution Approach 1:
The system extracts only the necessary eye gaze information from the driver's face without processing the complete scene data. This selective extraction approach maintains accurate driver state measurement while significantly reducing the computational resources and energy required for processing.
Solution Approach 2:
The system performs partial processing by focusing exclusively on eye gaze detection rather than comprehensive scene analysis. This partial action approach provides sufficient information for driver state estimation without the excessive resource consumption of full scene processing.
3Measurement precision
If the system uses divergence algorithms to calculate distance between expected and actual gaze distributions, then driver state estimation accuracy is improved, but device complexity increases
Solution Approach 1:
The patent introduces an intermediary neural network model that generates expected gaze distributions based on scene understanding. This intermediary component bridges the gap between simple eye tracking and complex scene analysis, enabling accurate driver state estimation through comparison of expected versus actual gaze patterns without requiring full explicit scene detection.
Data Source
AI summary
A system and method for estimation of a driver state based on eye gaze includes, capturing and sending, using an outward looking camera situated in a vehicle, a first video stream of surrounding environment to a neural controller. The neural controller, based on the first video stream, generates an expected gaze distribution. Using an inward looking camera situated in the vehicle, the camera captures and sends a second video stream of a face of a driver to an eye tracker controller, where based on the second video stream, the eye tracker controller extracts a plurality of gaze directions. A gaze distribution module generates, based on the plurality of gaze directions, an actual gaze distribution. A distance distribution controller, based on a difference between the expected gaze distribution and the actual gaze distribution, generates a distance measure where a determination is made that the distance measure exceeds a threshold.