Adaptive nametag training with exogenous inputs
First Claim
Patent Images
1. A method of speech recognition comprising:
- receiving an utterance at a vehicle telematics unit;
converting the utterance into at least one phoneme;
determining a confidence score based on a comparison between the at least one phoneme and a nametag phoneme; and
storing the at least one phoneme in association with the nametag phoneme based on the confidence score.
3 Assignments
0 Petitions
Accused Products
Abstract
A method of speech recognition includes receiving an utterance at a vehicle telematics unit. The utterance is converted into at least one phoneme. A confidence score is determined based on a comparison between the at least one phoneme and a nametag phoneme. The at least one phoneme is stored in association with the nametag phoneme based on the confidence score.
-
Citations
17 Claims
-
1. A method of speech recognition comprising:
-
receiving an utterance at a vehicle telematics unit;
converting the utterance into at least one phoneme;
determining a confidence score based on a comparison between the at least one phoneme and a nametag phoneme; and
storing the at least one phoneme in association with the nametag phoneme based on the confidence score. - View Dependent Claims (2, 3, 4, 5, 6, 7)
-
-
8. The method of claim 8 further comprising managing storage space for the alternative nametags.
-
9. A computer usable medium including a program for speech recognition comprising:
-
computer readable program code for receiving an utterance at a vehicle telematics unit;
computer readable program code for converting the utterance into at least one phoneme;
computer readable program code for determining a confidence score based on a comparison between the at least one phoneme and a nametag phoneme; and
computer readable program code for storing the at least one phoneme in association with the nametag phoneme based on the determined confidence score. - View Dependent Claims (10, 11, 12, 13, 14, 15, 16)
-
-
17. A speech recognition system comprising:
-
means for receiving an utterance at a vehicle telematics unit;
means for converting the utterance into at least one phoneme;
means for determining a confidence score based on a comparison between the at least one phoneme and a nametag phoneme; and
means for storing the at least one phoneme in association with the nametag phoneme based on the determined confidence score.
-
Specification