Multi-Microphone Speech Recognition Systems and Related Techniques

• US 20160358619A1
• Filed: 06/06/2015
• Published: 12/08/2016
• Est. Priority Date: 06/06/2015
• Status: Active Grant

• Current Assignee: Apple Inc.
• Original Assignee: Apple Inc.
• Inventors: Ramprashad, Sean A., Thornburg, Harvey D., Krishnaswamy, Arvindh, Lindahl, Aram M.
• Granted Patent: US 9,865,265 B2
• Time in Patent Office: Days
• US Class Current: 1/1
• CPC Class Codes: G10L 15/16 using artificial neural networks
• CPC Class Codes: G10L 15/20 Speech recognition techniques specially adapted for robustness in adverse environments, e.g. in noise, of stress induced speech G10L21/02 takes precedence
• CPC Class Codes: G10L 15/34 Adaptation of a single recogniser for parallel processing, e.g. by use of multiple processors or cloud computing
• CPC Class Codes: G10L 2015/022 Demisyllables, biphones or triphones being the recognition units
• CPC Class Codes: G10L 2021/02166 Microphone arrays; Beamforming

Abstract

A speech recognition system for resolving impaired utterances can have a speech recognition engine configured to receive a plurality of representations of an utterance and concurrently to determine a plurality of highest-likelihood transcription candidates corresponding to each respective representation of the utterance. The recognition system can also have a selector configured to determine a most-likely accurate transcription from among the transcription candidates. As but one example, the plurality of representations of the utterance can be acquired by a microphone array, and beamforming techniques can generate independent streams of the utterance across various look directions using output from the microphone array.

Claims

1. A speech recognition system for resolving far-field utterances, comprising:

a recognition engine configured to concurrently receive a plurality of representations of an utterance and to determine a highest-probability representation of the utterance; and

an utterance decoder configured to determine a most-likely transcription corresponding to the highest-probability representation of the utterance.

2. The speech recognition system according to claim 1, wherein the plurality of representations of the utterance comprises a concatenation of utterance representations and corresponding posterior probabilities.

3. The speech recognition system according to claim 1, wherein each of the utterance representations has an associated posterior probability, and wherein the speech recognition engine is configured to determine the highest-probability representation of the utterance based in part on a combination of the plurality of posterior probabilities corresponding to the utterance representations.

4. The speech recognition system according to claim 1, wherein the speech recognition engine more heavily weights utterance representations having relatively higher signal-to-noise ratio as compared to utterance representations having relatively lower signal-to-noise ratio.

9. The speech recognition system according to claim 1, further comprising a recognition parameter store, wherein the utterance decoder is configured to combine one or more recognition parameters from the recognition parameter store with the highest-probability representations of the utterance.

10. The speech recognition system according to claim 9, wherein the plurality of representations of the utterance comprises a plurality of acoustic features, each having a corresponding posterior probability, and wherein the speech recognition engine is configured to identify the acoustic features having a highest-probability of correctly representing the utterance, and wherein the utterance decoder is further configured to combine the one or more recognition parameters from the recognition parameter store with the highest probability acoustic features.

11. The speech recognition system of claim 10, wherein the recognition parameter store comprises an acoustic feature dictionary, a language model, or both.

12. The speech recognition system of claim 11, wherein the plurality of utterance representations comprises a plurality of phonemes and the acoustic feature dictionary comprises a phonetic dictionary.

13. The speech recognition system according to claim 1, wherein each representation of the utterance comprises one or more respective acoustic features and corresponding posterior probabilities, the system further comprising:

a hub configured to receive a stream of the acoustic features and corresponding posterior probabilities representative of the utterance from each of a plurality of acoustic feature extractors;

an aggregator configured to aggregate the plurality of streams and corresponding posterior probabilities; and

a communication connection configured to transmit the aggregated plurality of streams and corresponding posterior probabilities to the speech recognition engine configured to select from the concatenated plurality of streams those acoustic features most likely to accurately reflect the utterance.

14. The distributed speech recognition system according to claim 13, further comprising an appliance configured to extract acoustic features from an acoustic signal received by the appliance and to stream the extracted acoustic features to the hub.

15. The distributed speech recognition system according to claim 14, wherein the appliance comprises a first appliance, the speech recognition system further comprising a second appliance configured to extract acoustic features an acoustic signal received by the second appliance and to stream the extracted acoustic features to the hub.

16. The distributed speech recognition system according to claim 15, wherein the first appliance comprises a near-field acoustic-feature extractor, a far-field acoustic-feature extractor, or both.

17. The distributed speech recognition system according to claim 15, wherein the hub is configured to synchronize the plurality of received streams of acoustic features and associated posterior probabilities.

5. (canceled)

6. (canceled)

7. (canceled)

8. (canceled)

18. A speech-recognition method, comprising:

selecting a highest-probability representation of an utterance from a plurality of concurrently generated representations of the utterance; and

determining a most-likely transcription of the utterance in correspondence to the highest-probability representation of the utterance.

19. The speech recognition method according to claim 18, wherein each of the utterance representations has an associated posterior probability, and wherein the act of selecting the highest-probability representation comprises combining the plurality of posterior probabilities corresponding to the utterance representations.

20. The speech recognition method according to claim 19, wherein the act of selecting the highest-probability representation further comprises more heavily weighting utterance representations having relatively higher signal-to-noise ratio as compared to utterance representations having relatively lower signal-to-noise ratio.

21. (canceled)

22. (canceled)

23. (canceled)

24. The speech recognition method according to claim 23, further comprising combining one or more recognition parameters from a recognition parameter store with the highest-probability utterance representations.

25. A non-transitory, computer-readable media containing instructions that, when executed by a processor, cause a computing environment to perform a speech recognition method comprising:

selecting a highest-probability representation of an utterance from a plurality of concurrently generated representations of the utterance; and

determining a most-likely transcription of the utterance in correspondence to the highest-probability representation of the utterance.

26. The non-transitory, computer-readable media according to claim 25, wherein each of the utterance representations has an associated posterior probability, and wherein the act of selecting the highest-probability representation comprises combining the plurality of posterior probabilities corresponding to the utterance representations.

27. The non-transitory, computer-readable media according to claim 26, wherein the act of selecting the highest-probability representation further comprises more heavily weighting utterance representations having relatively higher signal-to-noise ratio as compared to utterance representations having relatively lower signal-to-noise ratio.

28. (canceled)

29. (canceled)

30. (canceled)

31. (canceled)