Automated speech recognition confidence classifier

Patent metadata

Field	Value
Publication number	US-11620992-B2
Application number	US-202117218964-A
Country	US
Kind code	B2
Filing date	Mar 31, 2021
Priority date	Apr 8, 2019
Publication date	Apr 4, 2023
Grant date	Apr 4, 2023

Abstract

A method of enhancing an automated speech recognition confidence classifier includes receiving a set of baseline confidence features from one or more decoded words, deriving word embedding confidence features from the baseline confidence features, joining the baseline confidence features with word embedding confidence features to create a feature vector, and executing the confidence classifier to generate a confidence score, wherein the confidence classifier is trained with a set of training examples having labeled features corresponding to the feature vector.

First claim

The invention claimed is:

1. A method of enhancing an automated speech recognition confidence classifier comprising:
receiving a set of baseline confidence features;
deriving embedding confidence features;
joining the baseline confidence features with the embedding confidence features to create a feature vector; and
executing the confidence classifier to generate a confidence score, wherein the confidence classifier is trained with a set of training examples corresponding to the feature vector.

2. The method of claim 1 wherein the embedding confidence features comprise character embeddings.

3. The method of claim 2 wherein the character embeddings comprise less than 26 embeddings comprising letters.

4. The method of claim 2 wherein the character embedding for a word comprises a vector having values for each letter consisting of the count of the number of each letter in the word.

5. The method of claim 1 wherein the embedding confidence features comprise phone embeddings.

6. The method of claim 5 wherein the phone embeddings comprise monophones selected from a dictionary comprising 40 or fewer monophones.

7. The method of claim 1 wherein the embedding confidence features comprise character embeddings and phone embeddings.

8. The method of claim 1 wherein the feature vector further comprises GLOVE embeddings.

9. The method of claim 1 wherein the confidence classifier is trained for word-level as well an utterance-level classification.

10. The method of claim 1 wherein the baseline features comprise two or more of acoustic-model scores, background-model scores, silence-model scores, noise-model scores, language model scores, and duration features.

11. The method of claim 1 wherein joining the baseline confidence features with the embedding confidence features comprising concatenating the baseline confidence features with the embedding confidence features.

12. The method of claim 1 wherein the set of baseline confidence features comprise a lattice of alternative word-sequences corresponding to an utterance.

13. The method of claim 1 wherein the set of baseline confidence features include acoustic model and language model scores.

14. The method of claim 1 wherein an acoustic model score is a function of acoustic state.

15. A machine-readable storage device having instructions for execution by a processor of a machine to cause the processor to perform operations to generate a confidence score for a word or utterance, the operations comprising:
receiving a set of baseline confidence features;
deriving embedding confidence features;
joining the baseline confidence features with the embedding confidence features to create a feature vector; and
executing the confidence classifier to generate a confidence score, wherein the confidence classifier is trained with a set of training examples corresponding to the feature vector.

16. The device of claim 14 wherein the embedding confidence features comprise phone embeddings including monophones selected from a dictionary comprising 40 or fewer monophones.

17. The device of claim 14 wherein the confidence classifier is trained for word-level as well an utterance-level classification and wherein the baseline features comprise two or more of acoustic-model scores, background-model scores, silence-model scores, noise-model scores, language model scores, and duration features.

18. The device of claim 14 wherein the set of baseline confidence features include acoustic model and language model scores.

19. A device comprising:
a processor; and
a memory device coupled to the processor and having a program stored thereon for execution by the processor to perform operations comprising:
receiving a set of baseline confidence features;
deriving embedding confidence features;
joining the baseline confidence features with the embedding confidence features to create a feature vector; and
executing the confidence classifier to generate a confidence score, wherein the confidence classifier is trained with a set of training examples corresponding to the feature vector.

20. The device of claim 18 wherein the confidence classifier is trained for word-level as well an utterance-level classification and wherein the baseline features comprise two or more of acoustic-model scores, background-model scores, silence-model scores, noise-model scores, language model scores, and duration features.

Assignees

• Microsoft Technology Licensing Llc

Inventors

• Kumar Kshitiz
• Anastasakos Anastasios
• Gong Yifan

Classifications

• G10L15/16

  using artificial neural networks · CPC title

• G10L15/02

  Feature extraction for speech recognition; Selection of recognition unit · CPC title

• G10L15/08 Primary

  Speech classification or search · CPC title

• G10L15/183 Primary

  using context dependencies, e.g. language models · CPC title

• G10L15/22

  Procedures used during a speech recognition process, e.g. man-machine dialogue · CPC title