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:
Sep. 08, 2026

Filed:

Dec. 18, 2023
Applicant:

Samsara Inc., San Francisco, CA (US);

Inventors:

Akshay Raj Dhamija, Campbell, CA (US);

Sven Eberhardt, Seattle, WA (US);

Yan Wang, Mercer Island, WA (US);

Phil Ammirato, Campbell, CA (US);

Brian Westphal, Livermore, CA (US);

William Nathan Hurst, Seattle, WA (US);

William Evan Welbourne, Seattle, WA (US);

Assignee:

Samsara Inc., San Francisco, CA (US);

Attorney:
Primary Examiner:
Int. Cl.
CPC ...
G06V 10/764 (2022.01); G06T 7/11 (2017.01); H04L 67/306 (2022.01);
U.S. Cl.
CPC ...
G06V 10/764 (2022.01); G06T 7/11 (2017.01); H04L 67/306 (2013.01);
Abstract

A system and method are described for enabling end users to create and refine object detection models using natural language and visual inputs. A user provides a prompt describing a target object. This prompt is processed to generate a feature vector representation. One or more images are also processed to detect object regions which are encoded into feature vectors. The prompt vector is compared against the image region vectors to identify relevant matching images. These images are displayed back to the user, along with bounding boxes indicating detected object regions. The prompt can be iteratively refined, allowing users to customize object detectors tailored to their use cases without needing machine learning expertise or model retraining.


Find Patent Forward Citations

Loading…