Patent · US Active

Sensor and method for generating training data for a machine learning classifier in a sensor

US12326917B2 · kind B2 · utility

0Cited by

1References

11Claims

0Family size

Assignee

Robert Bosch GmbH · DE

Inventors

Stefan Leidich · Rutesheim, DE
Andreas Frischen · Leonberg, DE
Stefan Weiss · Trostberg, DE

Key dates

Filing date	Sep 1, 2021
Grant date	Jun 10, 2025
Priority date	—
Expiry date	Apr 11, 2044

Classification

Technology area (CPC G)Physics
CPC primaryG06F2218/12
WIPO fieldComputer technology
WIPO sectorElectrical engineering

Abstract

A sensor and a method for generating training data for a machine learning classifier in a sensor. The method includes: detecting a first sensor signal and a second sensor signal with the aid of the sensor; ascertaining whether the machine learning classifier classifies the first sensor signal as a predefined user motion pattern; ascertaining whether a classification model classifies the second sensor signal as a predefined user input for controlling a device, the user input being assigned to the user motion pattern; ascertaining a training data vector for the machine learning classifier if the second sensor signal has been classified as the predefined user input, and if the first sensor signal has not been classified as the user motion pattern, ascertaining a training data vector for the machine learning classifier, the training data vector including the first sensor signal as well as the user motion pattern.

Source: USPTO / EPO open patent data. Objective bibliographic and citation counts.