SYSTEM AND METHOD FOR MANAGING MODELS FOR EMBEDDED SPEECH AND LANGUAGE PROCESSING

• US 20150120287A1
• Filed: 10/28/2013
• Published: 04/30/2015
• Est. Priority Date: 10/28/2013
• Status: Active Grant

• Current Assignee: Hyundai Motor Company (Hyundai Motor Group), Kia Corp.
• Original Assignee: AT&T Intellectual Property I LP (AT&T, Inc.)
• Inventors: STERN, Benjamin J., BOCCHIERI, Enrico Luigi, CONKIE, Alistair D., GIULIANELLI, Danilo
• Granted Patent: US 9,530,416 B2
• Time in Patent Office: Days
• US Class Current: 704/231
• CPC Class Codes: G10L 15/183 using context dependencies, e.g. language models
• CPC Class Codes: G10L 15/22 Procedures used during a speech recognition process, e.g. man-machine dialogue
• CPC Class Codes: G10L 15/30 Distributed recognition, e.g. in client-server systems, for mobile phones or network applications
• CPC Class Codes: G10L 2015/228 of application context

Abstract

Disclosed herein are systems, methods, and computer-readable storage devices for fetching speech processing models based on context changes in advance of speech requests using the speech processing models. An example local device configured to practice the method, having a local speech processor, and having access to remote speech models, detects a change in context. The change in context can be based on geographical location, language translation, speech in a different language, user language settings, installing or removing an app, and so forth. The local device can determine a speech processing model that is likely to be needed based on the change in context, and that is not stored on the local device. Independently of an explicit request to process speech, the local device can retrieve, from a remote server, the speech processing model for use on the mobile device.

Claims

1. A method comprising:

detecting, at a mobile device, a change in context;

determining a speech processing model that is likely to be needed based on the change in context, and that is not stored on the mobile device; and

independently of an explicit request to process speech, retrieving, from a remote server, the speech processing model for use on the mobile device.

2. The method of claim 1, further comprising:

calculating a likelihood score that the speech processing model will be needed; and

retrieving the speech processing model only when the likelihood score is above a threshold.

3. The method of claim 1, further comprising:

determining that a plurality of speech processing models are likely to be needed based on the change in context;

determining priority rankings for the plurality of speech processing models; and

retrieving the plurality of speech processing models from the remote server based on the priority rankings

4. The method of claim 1, wherein the change in context is based on one of geographical location, language translation, speech in a different language, user language settings, available local storage, or installing or removing an app on the mobile device.

5. The method of claim 1, further comprising:

comparing network characteristics to a priority for the speech processing model to determine when to retrieve the speech processing model.

6. The method of claim 1, further comprising:

determining that an existing speech processing model stored on the device is unlikely to be needed based on the change in context; and

removing the existing speech processing model from local storage on the device.

7. The method of claim 1, further comprising:

analyzing a local storage of the mobile device to determine whether the local storage has sufficient storage space to store the speech processing model;

upon determining that the local storage has insufficient space to store the speech processing model;

identifying, based on the change in context, speech processing models stored in the local storage that are less likely to be needed than the speech processing model; and

removing a sufficient number of the speech processing models from the local storage to free storage space in the local storage to accommodate the storage model.

12. The system of claim 1, the computer-readable storage medium further stores instructions which result in the method further comprising:

determining that the speech processing model stored on the mobile device is also below a threshold likelihood of other speech processing models stored on the mobile device.

13. The system of claim 1, the computer-readable storage medium further stores instructions which result in the method further comprising:

determining that a plurality of speech processing models are below the threshold likelihood of use based on the change in context;

determining priority rankings for the plurality of speech processing models; and

removing the plurality of speech processing models from the mobile device based on the priority rankings

14. The system of claim 13, the computer-readable storage medium further stores instructions which result in the method further comprising:

removing the plurality of speech processing models from the mobile device based on the priority rankings until a threshold amount of storage space on the mobile device is freed.

8. A system comprising:

a processor; and

a computer-readable storage medium storing instructions which, when executed by the processor, cause the processor to perform a method comprising;

detecting, at a mobile device, a change in context;

determining that a speech processing model stored on the mobile device is below a threshold likelihood of use based on the change in context; and

removing at least a portion of the speech processing model from the mobile device in response to the change in context.

9. The system of claim 8, wherein the change in context is based on one of unused storage space on the mobile device crossing a threshold, geographical location, language translation, speech in a different language, user language settings, or installing or removing an app on the mobile device.

10. The system of claim 8, wherein the threshold of likelihood of use is based on available storage space for storing speech processing models.

11. The system of claim 8, wherein the threshold of likelihood of use is based on a size of an expected speech processing model to be received based on the change in context.

15. A computer-readable storage device storing instructions which, when executed by a computing device, cause the computing device to perform a method comprising:

detecting, at a mobile device, a change in context;

determining a speech processing model that is likely to be needed based on the change in context, and that is not stored on the mobile device; and

independently of an explicit request to process speech, retrieving, from a remote server, the speech processing model for use on the mobile device.

16. The computer-readable storage device of claim 15, storing additional instructions which result in the method further comprising:

calculating a likelihood score that the speech processing model will be needed; and

retrieving the speech processing model only when the likelihood score is above a threshold.

17. The computer-readable storage device of claim 15, storing additional instructions which result in the method further comprising:

determining that a plurality of speech processing models are likely to be needed based on the change in context;

determining priority rankings for the plurality of speech processing models; and

retrieving the plurality of speech processing models from the remote server based on the priority rankings

18. The computer-readable storage device of claim 15, wherein the change in context is based on one of geographical location, language translation, speech in a different language, user language settings, or installing or removing an app on the mobile device.

19. The computer-readable storage device of claim 15, storing additional instructions which result in the method further comprising:

comparing network characteristics to a priority for the speech processing model to determine when to retrieve the speech processing model.

20. The computer-readable storage device of claim 15, storing additional instructions which result in the method further comprising:

analyzing a local storage of the mobile device to determine whether the local storage has sufficient storage space to store the speech processing model;

upon determining that the local storage has insufficient space to store the speech processing model;

identifying, based on the change in context, a least-needed speech processing model stored in the local storage; and

removing the least-needed speech processing model from the local storage.