IP Library › Granted Patent US 12,548,557
Granted Patent B2
US 12,548,557 · App. 18/173,699 · Granted Feb 10, 2026

Method for human speech processing

Inventors: Rémi Louis Clément Ponçot (Grenoble, FR); Abbas Ataya (Menlo Park, CA); Peter George Hartwell (Menlo Park, CA)
Assignee: TDK CORPORATION
G10L15/08G10L15/30G10L2015/088
View Patent ↗
Loading inventors, assignments & file history…
Monitor This Case
Get email alerts when status or documents change.
Order Certified Copies
Most orders are placed with the USPTO same day — all within 24 business hours.
Order via The Patent Place →
Pre-filled with this patent's details
Quick Facts
Patent No.
US 12,548,557
App. No.
18/173,699
Granted
Feb 10, 2026
Kind
B2
Abstract

In a method for human speech processing in an automatic speech recognition (ASR) system, human speech is received at a speech interface of the ASR system, wherein the ASR system comprises embedded componentry for onboard processing of the human speech and cloud-based componentry for remote processing of the human speech. A keyword is identified at the speech interface within a first portion of the human speech. Responsive to identifying the keyword, a second portion of the human speech is analyzed to identify at least one command, the second portion following the first portion. The at least one command is identified within the second portion of the human speech. The at least one command is selectively processed within at least one of the embedded componentry and the cloud-based componentry.

Claims (39)

1 . A method for human speech processing in an automatic speech recognition (ASR) system, the method comprising:

receiving human speech at a speech interface of the ASR system, wherein the ASR system comprises embedded componentry for onboard processing of the human speech and cloud-based componentry for remote processing of the human speech;

identifying a keyword at the speech interface within a first portion of the human speech, wherein the keyword is identified prior to completion of a last phoneme of the keyword;

responsive to identifying the keyword, analyzing a second portion of the human speech to identify at least one command, the second portion following the first portion;

identifying the at least one command within the second portion of the human speech;

forwarding the at least one command to the embedded componentry and the cloud-based componentry; and

concurrently processing the at least one command within the embedded componentry and the cloud-based componentry.

2 . The method of claim 1 , wherein the selectively processing the at least one command within at least one of the embedded componentry and the cloud-based componentry comprises:

determining whether the at least one command is executable at the embedded componentry;

provided the at least one command is capable of being processed at the embedded componentry, processing the at least one command at the embedded componentry and canceling processing of the at least one command at the cloud-based componentry; and

provided the at least one command is not capable of being processed at the embedded componentry, continuing processing the at least one command at the cloud-based componentry.

3 . The method of claim 2 , wherein the determining whether the at least one command is executable at the embedded componentry comprises:

determining whether the at least one command is within an inventory of commands capable of being processed by the embedded componentry.

4 . A non-transitory computer readable storage medium having computer readable program code stored thereon for causing a computer system to perform a method for human speech processing in an automatic speech recognition (ASR) system, the method comprising:

receiving human speech at a speech interface of the ASR system, wherein the ASR system comprises embedded componentry for onboard processing of the human speech and cloud-based componentry for remote processing of the human speech;

identifying a keyword at the speech interface within a first portion of the human speech, wherein the keyword is identified prior to completion of a last phoneme of the keyword;

responsive to identifying the keyword, analyzing a second portion of the human speech to identify at least one command, the second portion following the first portion;

identifying the at least one command within the second portion of the human speech;

forwarding the at least one command to the embedded componentry and the cloud-based componentry; and

concurrently processing the at least one command within the embedded componentry and the cloud-based componentry.

5 . The computer readable storage medium of claim 4 , wherein the selectively processing the at least one command within at least one of the embedded componentry and the cloud-based componentry comprises:

determining whether the at least one command is executable at the embedded componentry;

provided the at least one command is capable of being processed at the embedded componentry, processing the at least one command at the embedded componentry and canceling processing of the at least one command at the cloud-based componentry; and

provided the at least one command is not capable of being processed at the embedded componentry, continuing processing the at least one command at the cloud-based componentry.

6 . The computer readable storage medium of claim 5 , wherein the determining whether the at least one command is executable at the embedded componentry comprises:

determining whether the at least one command is within an inventory of commands capable of being processed by the embedded componentry.

7 . A computer system comprising:

a data storage unit; and

a processor coupled with the data storage unit, the processor configured to:

receive human speech at a speech interface of an automatic speech recognition (ASR) system, wherein the ASR system comprises embedded componentry for onboard processing of the human speech and cloud-based componentry for remote processing of the human speech;

identify a keyword at the speech interface within a first portion of the human speech, wherein the keyword is identified prior to completion of a last phoneme of the keyword;

analyze a second portion of the human speech to identify at least one command responsive to identifying the keyword, the second portion following the first portion;

identify the at least one command within the second portion of the human speech;

forward the at least one command to the embedded componentry and the cloud-based componentry; and

concurrently process the at least one command within the embedded componentry and the cloud-based componentry.

8 . The computer system of claim 7 , wherein the processor is further configured to:

determine whether the at least one command is executable at the embedded componentry;

process the at least one command at the embedded componentry provided the at least one command is capable of being processed at the embedded componentry; and

process the at least one command at the cloud-based componentry provided the at least one command is not capable of being processed at the embedded componentry.

Assignments (4)
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Jun 13, 2023
From: PONÇOT, RÉMI LOUIS CLÉMENT
To: MOVEA SAS
Reel/Frame 063939/0131 →
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Jun 13, 2023
From: MOVEA SAS
To: INVENSENSE, INC.
Reel/Frame 063939/0795 →
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Jun 13, 2023
From: ATAYA, ABBAS; HARTWELL, PETER GEORGE
To: INVENSENSE, INC.
Reel/Frame 063939/0817 →
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Jun 13, 2023
From: INVENSENSE, INC.
To: TDK CORPORATION
Reel/Frame 063939/0876 →
Continuity (2)
Provisional Application 63268431 · Feb 23, 2022
Related Publication 20230267919A1 · Aug 24, 2023
References Cited (6)
US 9070367B1 · Hoffmeister · 2015 [cited by examiner]
US 12190875B1 · Fidler · 2025 [cited by examiner]
US 20130085753A1 · Bringert · 2013 [cited by examiner]
US 20150006166A1 · Schmidt · 2015 [cited by examiner]
US 20160162469A1 · Santos · 2016 [cited by examiner]
US 20220036896A1 · Elkhatib · 2022 [cited by examiner]