Gaze-Guided Vehicle Occupancy Estimation for Distant Objects
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Occupancy estimation systems face reduced accuracy for objects far away from vehicles due to low camera resolution, leading to poor signal-to-noise ratios and unclear image data.
Innovation Solution
An occupancy estimation system that utilizes gaze estimation to improve object detection by tracking the gaze of a vehicle occupant, generating a gaze saliency map, and updating occupancy probabilities based on gaze direction and identified objects of interest, employing a gaze prediction model and Gaussian functions to enhance image clarity and accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If camera resolution is used for occupancy estimation of distant objects, then the system can detect objects at a distance, but the accuracy is reduced due to low resolution and poor signal-to-noise ratio
Solution Approach 1:
The patent introduces an occupancy estimation network as an intermediary that processes camera images to generate occupancy probability maps. This network acts as a mediator between the low-quality camera input and the required high-precision occupancy information, using learned features to compensate for the poor signal-to-noise ratio in distant object detection
Solution Approach 2:
The system transforms the problem from direct image analysis to occupancy probability estimation by changing the parameter space. Instead of relying on pixel-level details from low-resolution images, the system estimates occupancy probabilities at voxel level, transforming spatial and probabilistic parameters to achieve better measurement precision
2Measurement precision
If gaze tracking is added to improve occupancy estimation, then accuracy for distant objects improves, but system complexity increases
Solution Approach 1:
The camera system serves multiple functions: it captures both the external scene for occupancy estimation and the occupant's gaze direction. By making the camera system multi-functional, the patent avoids adding separate sensors while still incorporating gaze information to improve detection accuracy
Solution Approach 2:
The system uses gaze direction as feedback to update occupancy probabilities. The occupant's gaze provides information about objects of interest, and this feedback is integrated into the occupancy estimation network to refine the probability estimates for distant objects, improving accuracy without requiring additional hardware
Data Source
AI summary
A computer-implemented method when executed by data processing hardware causes the data processing hardware to perform operations. The operations include generating, at an occupancy estimation network, an occupancy probability at one or more voxels, predicting, via a gaze prediction model, a gaze direction of a gaze prediction, and identifying, based on the predicted gaze direction, an object of interest. The operations also include generating, via an occupancy estimation application, a gaze saliency map based on the gaze direction and identified object of interest, and updating, based on the determined gaze direction and the gaze saliency map, the occupancy probability of the one or more voxels.