Autonomous Vehicle Object Classification via Driver Gaze
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Autonomous vehicles face significant processing power challenges in real-time sensor data processing, particularly when additional sensors are integrated, leading to increased computational demands.
Innovation Solution
A system that includes an image capture device, a driver gaze capture device, and a processor, which classifies objects in the vehicle's environment using deep neural networks and builds a decision-making database based on driver gaze information, allocating computational resources efficiently by distinguishing between critical and non-critical objects.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If additional sensors are integrated to capture more environmental data, then the quality and completeness of autonomous vehicle perception is improved, but the processing power and computational demands increase significantly
Solution Approach 1:
The patent segments objects in the environment into different priority levels (first priority objects and second priority objects) based on their relevance to vehicle operation. This segmentation allows the system to process only critical objects with high computational resources while using reduced processing for less critical objects, thereby resolving the contradiction between comprehensive perception and processing power requirements.
2Reliability
If all detected objects are processed with equal computational resources, then comprehensive environmental awareness is achieved, but system resource usage increases unnecessarily
Solution Approach 1:
The patent applies local quality by assigning different processing qualities to different objects based on their priority classification. First priority objects receive full computational processing while second priority objects receive reduced processing, optimizing system resource usage while maintaining reliable environmental awareness for critical elements.
3Measurement precision
If deep neural network machine learning is used to classify all objects, then classification accuracy is improved, but computational time and processing power increase
Solution Approach 1:
The patent applies partial action by using deep neural network machine learning only for classifying first priority objects that require high accuracy, while using alternative classification methods for second priority objects. This approach maintains high classification accuracy for critical objects while reducing overall computational time and processing power consumption.
Data Source
AI summary
Systems and techniques for autonomous vehicle decision making may include training an autonomous vehicle decision making database by capturing an image including a first training object and a second training object during a training phase. The first training object may be classified as a first class and the second training object may be classified as a second class based on a driver gaze location associated with a driver of the vehicle. The database may be built based on classification of the first training object and the second training object. The autonomous vehicle decision making database may be utilized to classify a first object as a first class and a second object as a second class during an operation phase. A processor may perform a first computation associated with the first object based on the classification of the first object and the classification of the second object.