SYSTEMS AND METHODS FOR SPEECH PROCESSING COMPRISING ADJUSTMENT OF HIGH FREQUENCY ATTACK AND RELEASE TIMES

Patent №

US 10,020,006

Granted

2018-07-10

Filed 2016

Owner

YOBE, INC.

Lab

AI components

2

nlp · speech

Assignment

Recorded

Dataset

AIPD

2023_r1 edition

Application

15243688

Systems and methods described herein modify audio content on an electronic device. Embodiments can be configured to detect a mode of the electronic device to determine whether the device is in a telephone mode; receive a speech signal from a speech source while the device is in the telephone mode; and process the speech signal to improve the perceived quality of the speech at a recipient when the electronic device is in a telephone mode; wherein processing the speech signal to improve the perceived quality of the speech comprises, decreasing the signal level of audio content outside of a determined frequency band relative to the signal level of the audio content within the determined frequency band; and wherein the determined frequency band is a frequency band associated a vocal range of the anticipated speech content. The method further includes adjusting high frequency sounds such as attack and release times of the speech signal based on sound events within the speech signal.

Natural languageSpeechG10L 21/0364G10L 21/0208G10L 25/81G10L 19/00H03G 5/005H03G 5/22

AI classification

Speech1.00
Natural language1.00
Vision0.02
AI hardware0.02
Planning0.01
Evolutionary computation0.00
Machine learning0.00
Knowledge representation0.00

Ownership

YOBE, INC.

assignment · 430800761

Assignors

FAIREY, JAMES CHRISTOPHER

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

© 2026 NYSGPT2525 LLC