MIDDLE-END SOLUTION TO ROBUST SPEECH RECOGNITION

Patent №

US 7,516,069

Granted

2009-04-07

Filed 2004

Owner

TEXAS INSTRUMENTS INCORPORATED

Lab

AI components

3

ml · nlp · speech

Assignment

Recorded

Dataset

AIPD

2023_r1 edition

Application

10823059

A method for performing time and frequency Signal-to-Noise Ratio (SNR) dependent weighting in speech recognition is described that includes for each period t estimating the SNR to get time and frequency SNR information ηt,f; calculating the time and frequency weighting to get γtf; performing the back and forth weighted time varying DCT transformation matrix computation MGtM−1 to get Tt; providing the transformation matrix Tt and the original MFCC feature ot that contains the information about the SNR to a recognizer including the Viterbi decoding; and performing weighted Viterbi recognition bj(ot).

AI classification

Speech1.00
Machine learning0.98
Natural language0.95
Vision0.13
AI hardware0.02
Evolutionary computation0.00
Knowledge representation0.00
Planning0.00

Ownership

TEXAS INSTRUMENTS INCORPORATED

assignment · 155450638

Assignors

BERNARD, ALEXIS P., GONG, YIFAN

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

© 2026 NYSGPT2525 LLC