The patent badge is an abbreviated version of the USPTO patent document. The patent badge does contain a link to the full patent document.

The patent badge is an abbreviated version of the USPTO patent document. The patent badge covers the following: Patent number, Date patent was issued, Date patent was filed, Title of the patent, Applicant, Inventor, Assignee, Attorney firm, Primary examiner, Assistant examiner, CPCs, and Abstract. The patent badge does contain a link to the full patent document (in Adobe Acrobat format, aka pdf). To download or print any patent click here.

Date of Patent:
Apr. 29, 2025

Filed:

Dec. 09, 2020
Applicant:

Nvidia Corporation, Santa Clara, CA (US);

Inventors:

Trung Pham, Santa Clara, CA (US);

Berta Rodriguez Hervas, San Francisco, CA (US);

Minwoo Park, Saratoga, CA (US);

David Nister, Bellevue, WA (US);

Neda Cvijetic, East Palo Alto, CA (US);

Assignee:

NVIDIA Corporation, Santa Clara, CA (US);

Attorney:
Primary Examiner:
Assistant Examiner:
Int. Cl.
CPC ...
B60W 30/18 (2011.12); G06N 3/04 (2022.12); G06N 3/08 (2022.12); G06T 7/33 (2016.12); G06V 10/764 (2021.12); G06V 10/82 (2021.12); G06V 20/56 (2021.12); G06V 20/64 (2021.12); G06N 3/045 (2022.12);
U.S. Cl.
CPC ...
B60W 30/18154 (2012.12); G06N 3/04 (2012.12); G06N 3/08 (2012.12); G06T 7/33 (2016.12); G06V 10/764 (2021.12); G06V 10/82 (2021.12); G06V 20/56 (2021.12); G06V 20/647 (2021.12); B60W 2520/06 (2012.12); G06T 2207/30252 (2012.12);
Abstract

In various examples, a three-dimensional (3D) intersection structure may be predicted using a deep neural network (DNN) based on processing two-dimensional (2D) input data. To train the DNN to accurately predict 3D intersection structures from 2D inputs, the DNN may be trained using a first loss function that compares 3D outputs of the DNN—after conversion to 2D space—to 2D ground truth data and a second loss function that analyzes the 3D predictions of the DNN in view of one or more geometric constraints—e.g., geometric knowledge of intersections may be used to penalize predictions of the DNN that do not align with known intersection and/or road structure geometries. As such, live perception of an autonomous or semi-autonomous vehicle may be used by the DNN to detect 3D locations of intersection structures from 2D inputs.


Find Patent Forward Citations

Loading…