METHOD FOR SUPPRESSING SPEAKER ACTIVATION IN A PORTABLE COMMUNICATION DEVICE OPERATED IN A SPEAKPHONE MODE

Patent №

US 6,138,040

Granted

2000-10-24

Filed 1998

Owner

MOTOROLA, INC.

Lab

AI components

1

speech

Assignment

Recorded

Dataset

AIPD

2023_r1 edition

Application

09127692

In a portable communication device (100) operated in a speakerphone mode, when no speech is present in either the outbound signal received from a communication system, the speaker (114) of the communication device is muted and the microphone (120) activated. When speech is detected in an inbound signal (202, 204, 206), and speech is also subsequently detected in the outbound signal (214), the speaker is left inactivated and the microphone kept in an activated state. To determine if speech is present in the inbound signal, the inbound signal is voided, providing a succession of frames (206), each frame having a frame energy parameter and a background noise parameter (208). If the frame energy of a given frame sufficiently exceeds the an average background noise value for the given frame, then speech is declared present. To smooth this half duplex operation, a timer may be used so that brief periods of silence or fricative parts of speech do not result in a revocation of the declaration of speech in the inbound signal.

AI classification

Speech1.00
Planning0.28
Natural language0.08
Vision0.06
Knowledge representation0.02
AI hardware0.00
Machine learning0.00
Evolutionary computation0.00

Ownership

MOTOROLA, INC.

assignment · 93660433

Assignors

NICHOLLS, RICHARD BRENT, KOENIGSMAN, SCOTT DAVID

On an employer assignment, the assignors are typically the inventors.

© 2026 NYSGPT2525 LLC