System and method for standardized speech recognition infrastructure
View Patent ↗Disclosed herein are systems, methods, and computer-readable storage media for selecting a speech recognition model in a standardized speech recognition infrastructure. The system receives speech from a user, and if a user-specific supervised speech model associated with the user is available, retrieves the supervised speech model. If the user-specific supervised speech model is unavailable and if an unsupervised speech model is available, the system retrieves the unsupervised speech model. If the user-specific supervised speech model and the unsupervised speech model are unavailable, the system retrieves a generic speech model associated with the user. Next the system recognizes the received speech from the user with the retrieved model. In one embodiment, the system trains a speech recognition model in a standardized speech recognition infrastructure. In another embodiment, the system handshakes with a remote application in a standardized speech recognition infrastructure.
1. A computer-implemented method comprising:
initiating a voice call with a remote application;
determining, by processor-executable instructions executed by a processor, whether the remote application can apply standardized speech recognition models;
when the remote application can apply the standardized speech recognition models:
transmitting a speaker specific model to the remote application; and
the processor-executable instructions when executed by the processor instructing the remote application to recognize speech with the transmitted speaker specific model; and
when the remote application can not apply standardized speech recognition models:
the processor-executable instructions when executed by the processor determining whether the remote application can apply transformations of a standard speech recognition model.
2. The computer-implemented method of claim 1 , further comprising:
when the remote application can apply transformations of the standard speech recognition model:
transmitting transformations of a generic model to the remote application; and
instructing the remote application to recognize speech based on the transmitted transformations; and
when the remote application can not apply transformations of the standard speech recognition model, instructing the remote application to recognize speech with the generic model.
3. The computer-implemented method of claim 1 , wherein the speaker specific model is generated on a first device and the transformations are generated on a second device distinct from the first device.
4. The computer-implemented method of claim 1 , further comprising instructing the remote application to perform self-adaptation based on one of the transmitted speaker specific model and the transformations.
5. The computer-implemented method of claim 1 , wherein the transmitting occurs using an edge device in a communications network.
6. The computer-implemented method of claim 1 , wherein the remote application is a game, and wherein the game identifies voice commands in recognized speech and controls elements of the game based on the voice commands.
7. The computer-implemented method of claim 1 , wherein the generic model and the speaker specific model are publicly available.
8. The computer-implemented method of claim 1 , wherein a network client authenticates the speech before instructing the remote application to recognize the speech.
9. A system comprising:
a processor; and
a computer-readable storage medium having instructions stored which, when executed by the processor, cause the processor to perform operations comprising:
initiating a voice call with a remote application;
determining if the remote application can apply standardized speech recognition models;
when the remote application can apply the standardized speech recognition models:
transmitting a speaker specific model to the remote application, and
instructing the remote application to recognize speech with the transmitted model;
when the remote application can not apply standardized speech recognition models:
determining if the remote application can apply transformations of a standard speech recognition model.
10. The system of claim 9 , the computer-readable storage medium having additional instructions stored which result in operations comprising:
when the remote application can apply transformations of the standard speech recognition model:
transmitting transformations of a generic model to the remote application, and
instructing the remote application to recognize speech based on the transmitted transformations; and
when the remote application can not apply transformations of the standard speech recognition model, instructing the remote application to recognize speech with the generic model.
11. The system of claim 9 , wherein the speaker specific model is generated on a first device and the transformations are generated on a second device distinct from the first device.
12. The system of claim 9 , the computer-readable storage medium having additional instructions stored which result in operations comprising instructing the remote application to perform self-adaptation based on one of the transmitted speaker specific model and the transformations.
13. The system of claim 9 , wherein the transmitting occurs using an edge device in a communications network.
14. The system of claim 9 , wherein the remote application is a game, and wherein the game identifies voice commands in recognized speech and controls elements of the game based on the voice commands.
15. The system of claim 9 , wherein the generic model and the speaker specific model are publicly available.
16. The system of claim 9 , wherein a network client authenticates the speech before instructing the remote application to recognize the speech.
17. A computer-readable storage device having instructions stored which, when executed by a computing device, cause the computing device to perform operations comprising:
initiating a voice call with a remote application;
determining if the remote application can apply standardized speech recognition models;
when the remote application can apply the standardized speech recognition models:
transmitting a speaker specific model to the remote application, and
instructing the remote application to recognize speech with the transmitted model; and
when the remote application can not apply standardized speech recognition models:
determining if the remote application can apply transformations of a standard speech recognition model.
18. The computer-readable storage device of claim 17 , having additional instructions stored which result in operations comprising:
when the remote application can apply transformations of the standard speech recognition model:
transmitting transformations of a generic model to the remote application, and
instructing the remote application to recognize speech based on the transmitted transformations; and
when the remote application can not apply transformations of the standard speech recognition model, instructing the remote application to recognize speech with the generic model.
19. The computer-readable storage device of claim 17 , wherein the speaker specific model is generated on a first device and the transformations are generated on a second device distinct from the first device.
20. The computer-readable storage device of claim 17 , having additional instructions stored which result in operations comprising instructing the remote application to perform self-adaptation based on one of the transmitted speaker specific model and the transformations.