Solution Study: Machine Learning & Physics-Based Radar Simulation in Autonomous Vehicles
This presentation will illustrate how ML algorithms, such as convolutional neural networks and reinforcement neural networks, can benefit from physics-based radar simulation for autonomous driving applications. Using Ansys AVxcelerate Sensors and Ansys optiSlang, we will demonstrate the feasibility of using synthetic data as a source for ML training in radar based ADAS applications, and automatically generate labeled synthetic radar data for ML applications.
- Physics-based radar simulation in real time
- Automatically generate labeled radar data for machine learning applications
- Open API for radar integration with 3rd party tools
- Scenario variation for creating heterogenous training data sets