Driver Gaze Time-Window Analysis for Intentional Action Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing driving analysis technologies struggle to accurately differentiate between a driver's intentional actions and health deterioration or inattentiveness, often misidentifying intentional gaze deviations as signs of inattentiveness.
Innovation Solution
A driving analysis device that generates a time series of detection elements including line-of-sight direction and eye state, extracts feature quantities from this data, and analyzes the driver's condition using a function that takes these feature quantities as input, allowing for precise determination of intentional actions without focusing solely on the front direction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the system focuses solely on detecting gaze at the road front to evaluate driver inattentiveness, then the measurement process is simple, but the measurement precision deteriorates because intentional gaze deviations are misidentified as inattentiveness
Solution Approach 1:
The patent segments the driver's behavior analysis into multiple independent detection elements (line-of-sight direction, eye open/closed state, head orientation) and processes them separately through time-series analysis. This segmentation allows the system to capture nuanced behavioral patterns without requiring a single complex detection model, thereby improving measurement precision while managing system complexity.
Solution Approach 2:
The patent introduces dynamic time-series analysis with configurable time windows to evaluate driver behavior patterns over time rather than static snapshots. This dynamic approach allows the system to distinguish between intentional temporary gaze deviations and sustained inattentiveness, improving measurement precision by capturing the temporal dynamics of driver behavior.
2Reliability
If the system uses simple PRC calculation to evaluate driver inattentiveness, then the analysis method is easy to operate, but the reliability deteriorates due to misidentification of intentional actions as inattentiveness
Solution Approach 1:
The patent introduces an intermediary layer of feature quantity extraction that processes raw detection elements through time-windowed analysis before final evaluation. This intermediary processing layer transforms simple gaze data into meaningful behavioral patterns, improving reliability by filtering out false positives from intentional actions while maintaining operational ease through automated processing.
Solution Approach 2:
The patent implements feedback mechanisms where the system continuously monitors driver behavior patterns and adjusts evaluation based on temporal patterns. By analyzing sequences of detection elements over time windows, the system provides feedback that distinguishes between intentional temporary deviations and genuine inattentiveness, thereby improving reliability without significantly complicating operation.
3Measurement precision
If the system analyzes multiple detection elements over time windows, then the accuracy of distinguishing intentional actions from inattentiveness improves, but the loss of information increases due to complex data processing
Solution Approach 1:
The patent extracts only the essential feature quantities from the time-series detection elements that are relevant to distinguishing intentional actions from inattentiveness. By selectively extracting meaningful patterns (such as duration of gaze deviations, frequency of eye closures) rather than processing all raw data, the system maintains high measurement precision while minimizing information loss through focused feature extraction.
Data Source
AI summary
In a driving analysis device, an information generation unit generates a time series of detection elements including a line-of-sight direction indicating to which of a plurality of preset viewable areas a line of sight of a driver driving a vehicle is oriented and an open/closed state of driver's eyes. An acquisition unit acquires evaluation data from the time series of the detection elements using a time window having a preset time width. An extraction unit extracts a plurality of feature quantities including at least a result of summation of appearance frequencies with respect to each of the detection elements from the evaluation data. An analysis unit analyzes a driver's tendency using a function that receives, as input, at least a part of the feature quantities extracted by the extraction unit.