This project builds a unified neural network model that simultaneously performs positioning and Doppler estimation using imaging radar data. Through a self-supervised machine learning methodology, the project aims to significantly enhance perception performance compared to traditional methods.
CAM Pathfinder - Feasibility Studies
Unified Neural Net - Radar
Introducing the Unified Neural-Net Radar Project, studying the feasibility of developing a scalable and automated training network for imaging radar
About the project
Project objectives
Objectives for this project include the following:
- Unified Neural Network Model: Development of a robust model capable of simultaneous positioning and Doppler estimation using imaging radar data.
- Enhanced Perception Performance: Achieving superior object detection and classification.
- Benchmarking Reports: Evaluations comparing UNN model’s performance against existing solutions using industry-standard metrics.
- Demonstrator MVP: A software-based MVP for potential customers.
Project consortium
-
Radareye
-
Ensilica
Unified Neural Net – Radar is funded by UK Government, as part of the CAM Pathfinder Programme. CAM Pathfinder is delivered by DBT’s Connected and Automated Mobility (CAM) team, in partnership with Innovate UK and Zenzic
{{ teamMember.name }}
{{ teamMember.title }}