Neural Network Gaze Mapping for 3D Object Identification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional gaze determination systems are unable to pinpoint the specific object a subject is looking at, limiting their functionality in applications such as in-vehicle systems where precise interactions are necessary.
Innovation Solution
A regression-based machine learning model determines gaze direction using image data, projecting gaze vectors onto three-dimensional maps of surfaces to identify the object of interest, allowing systems to adapt to various environments by updating spatial regions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional gaze determination systems are used, then gaze direction can be determined generally, but the specific object being looked at cannot be pinpointed
Solution Approach 1:
The system segments the environment into multiple three-dimensional spatial regions or surfaces, allowing precise identification of which specific region the gaze vector intersects. This segmentation enables the system to distinguish between different objects or areas within the environment based on where the gaze vector points, thereby pinpointing the specific object being looked at.
Solution Approach 2:
The system transitions from two-dimensional gaze direction estimation to three-dimensional spatial region mapping by projecting gaze vectors onto three-dimensional maps of surfaces. This dimensional extension allows the system to determine not just the general direction of gaze but also the specific three-dimensional object or region being viewed, enhancing measurement precision without requiring complete system redesign.
2Adaptability or versatility
If the system is designed to work in arbitrary environments, then versatility is improved, but the complexity of defining spatial regions increases
Solution Approach 1:
The system uses a universal approach where the same three-dimensional spatial region definition method can be applied across different environments (vehicle interiors, rooms, etc.). By defining spatial regions in a standardized three-dimensional manner, the system can adapt to various environments without requiring environment-specific customization of the core processing logic, thereby achieving versatility while managing complexity through standardization.
Data Source
AI summary
Systems and methods for determining the gaze direction of a subject and projecting this gaze direction onto specific regions of an arbitrary three-dimensional geometry. In an exemplary embodiment, gaze direction may be determined by a regression-based machine learning model. The determined gaze direction is then projected onto a three-dimensional map or set of surfaces that may represent any desired object or system. Maps may represent any three-dimensional layout or geometry, whether actual or virtual. Gaze vectors can thus be used to determine the object of gaze within any environment. Systems can also readily and efficiently adapt for use in different environments by retrieving a different set of surfaces or regions for each environment.