Autonomous Vehicle Object Prediction via Machine Learning Models
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Autonomous vehicles face challenges in accurately predicting the future locations of objects in their environment, which is crucial for generating appropriate motion plans but is hindered by the limitations of existing prediction systems in terms of speed, quality, and accuracy.
Innovation Solution
A computer system that includes a scenario generation system and a scenario development system, utilizing machine-learned models such as static object classifiers, goal scoring models, and trajectory development models to generate and predict trajectories for objects based on their current and past states, as well as environmental data, enabling improved prediction of future object locations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional prediction systems are used to predict future object locations, then the system complexity is low, but the prediction accuracy and quality are insufficient
Solution Approach 1:
The prediction system is divided into multiple specialized machine-learned models including static object classifiers, goal scoring models, and trajectory development models. Each model handles a specific aspect of prediction, allowing the system to achieve high accuracy through modular specialization rather than a single complex monolithic system.
Solution Approach 2:
The system introduces intermediate processing layers including scenario generation and scenario development stages that bridge raw sensor data and final predictions. These intermediary processing steps enable complex reasoning about object goals and trajectories while maintaining system organization and manageability.
2Measurement precision
If complex machine learning models are deployed to improve prediction quality, then the prediction accuracy increases, but the processing speed decreases
Solution Approach 1:
By segmenting the prediction task into parallel specialized models (static object classifier, goal scoring model, trajectory development model), the system can process different aspects of prediction simultaneously, improving overall throughput while maintaining high accuracy through specialized processing for each aspect.
Solution Approach 2:
The system performs preliminary classification of objects as static or dynamic using the static object classifier before applying more computationally intensive trajectory prediction. This preliminary action filters out objects that don't require complex prediction, reducing overall processing load while maintaining accuracy for objects that do require detailed prediction.
Data Source
AI summary
The present disclosure provides systems and methods for predicting the future locations of objects that are perceived by autonomous vehicles. An autonomous vehicle can include a prediction system that, for each object perceived by the autonomous vehicle, generates one or more potential goals, selects one or more of the potential goals, and develops one or more trajectories by which the object can achieve the one or more selected goals. The prediction systems and methods described herein can include or leverage one or more machine-learned models that assist in predicting the future locations of the objects. As an example, in some implementations, the prediction system can include a machine-learned static object classifier, a machine-learned goal scoring model, a machine-learned trajectory development model, a machine-learned ballistic quality classifier, and/or other machine-learned models. The use of machine-learned models can improve the speed, quality, and/or accuracy of the generated predictions.