/Patent Tracker
NvidiaPatent applicationCars & transportInterest 7/10

Nvidia patents method combining camera and radar data to map 3D space

Nvidia describes a system that merges data from cameras and radar sensors to create a detailed 3D map of the surroundings. It uses neural networks to process these combined signals and identify which parts of the environment are occupied by objects.

Why it mattersThis approach helps self-driving systems detect objects more reliably by using two different types of sensors together, which is a key step for safer autonomous navigation.

In the filing’s own words

In various examples, multi-modal fusion networks for occupancy detection in autonomous and/or semi-autonomous systems and applications is described herein. Systems and methods are disclosed that use a multi-modal fusion network to perform object detection based on image data obtained using one or more image sensors along with RADAR data obtained using one or more RADAR sensors of a machine. For instance, the image data may be processed to generate one or more first features and the RADAR data may be processed to generate one or more second features. The first feature(s) may then be combined (e.g., fused, etc.) with the second feature(s) to generate one or more fused features. Additionally, the fused feature(s) may be processed using one or more neural networks to determine information associated with objects located within an environment, such as three-dimensional (3D) occupancy detections and/or classifications associated with the 3D occupancy detections.

What exactly they claim (claim 1)

1. A method comprising: generating, using one or more first encoders and based at least on image data obtained using one or more image sensors of a machine, one or more first Bird's Eye View (BEV) features associated with the image data; generating, using one or more second encoders and based at least on RADAR data obtained using one or more RADAR sensors of the machine, one or more second BEV features associated with the RADAR data; generating, using one or more neural networks and based at least on fusing the one or more first BEV features with the one or more second BEV features, output data indicating whether one or more voxels associated with an environment are occupied; and causing, based at least on the output data, the machine to perform one or more planning, control, or navigation operations.

Facts

Company
Nvidia (NVIDIA Corporation)
Official title
SENSOR FUSION FOR OBJECT DETECTION IN AUTONOMOUS AND SEMI-AUTONOMOUS SYSTEMS AND APPLICATIONS
Publication number
US20260279030A1
Application number
19342294
Filed
Sep 26, 2025
Published
Sep 17, 2026
Inventors
14
Claims
20

Classified under

G06V 10/806 Physics & computingG01S 13/89 Physics & computingG06V 10/40 Physics & computingG06V 10/764 Physics & computingG06V 20/58 Physics & computing
Read the full filing (official PDF) ↗ Open in the Patent Tracker

More from Nvidia in Cars & transport

Summaries are simplified from the filing’s own abstract and first claim. A patent filing is not a product announcement.
Patent data and drawings: U.S. Patent and Trademark Office (USPTO). Nebula is not affiliated with or endorsed by the USPTO.