Context-sensitive dynamic update of voice to text model in a voice-enabled electronic device

1. A method implemented by one or more processors of a client device, the method comprising:

generating, using a locally stored list of relevant context sensitive entities, voice to text model update data for a context sensitive parameter that is associated with the relevant context sensitive entities, wherein the voice to text model update data comprises decoding paths for the relevant context sensitive entities;

subsequent to generating the voice to text model update data:

receiving a voice input with a voice-enabled electronic device, the voice input including an original request that includes first and second portions, the second portion including a first context sensitive entity among the relevant context sensitive entities; and

in the voice-enabled electronic device, and responsive to receiving the first portion of the voice input:

performing local processing of the first portion of the voice input;

determining during the local processing that the first portion is associated with the context sensitive parameter;

in response to determining that the first portion is associated with the context sensitive parameter:

dynamically updating a local voice to text model, used by the voice-enabled electronic device, using the locally generated voice to text model update data, wherein dynamically updating the local voice to text model facilitates recognition of the first context sensitive entity in performing local processing of the second portion of the voice input;

generating, utilizing the dynamically updated local voice to text model, a recognition of the second portion of the voice input; and

causing performance of a voice action that is based on the recognition of the second portion of the voice input.

2. The method of claim 1 , wherein performing the local processing includes:

converting a digital audio signal of the voice input to text using a streaming voice to text module of the voice-enabled electronic device, wherein the streaming voice to text module dynamically generates a plurality of text tokens from the digital audio signal.

3. The method of claim 1 , wherein dynamically updating the local voice to text model comprises augmenting the local voice to text model using the decoding paths.

4. The method of claim 1 , wherein determining during the local processing that the first portion is associated with the context sensitive parameter includes determining that the first portion includes a request to play a media item, wherein the context sensitive parameter includes a media item identifier for use in identifying the media item, and wherein the plurality of context sensitive entities identify a plurality of media items playable by the voice-enabled electronic device.

5. The method of claim 1 , wherein determining during the local processing that the first portion is associated with the context sensitive parameter includes determining that the first portion includes a request to communicate with a contact, wherein the context sensitive parameter includes a contact identifier for use in initiating a communication with the contact, and wherein the plurality of context sensitive entities identify a plurality of contacts accessible by the voice-enabled electronic device.

6. The method of claim 1 , wherein the context sensitive parameter is a location-dependent parameter, and wherein the plurality of context sensitive entities identify a plurality of points of interest disposed in proximity to a predetermined location.

7. The method of claim 6 , wherein the predetermined location comprises a current location of the voice-enabled electronic device, the method further comprising, in response to a determination that the first portion is associated with the context sensitive parameter, communicating the current location to an online service and prefetching from the online service voice to text model update data associated with the plurality of context sensitive entities.

8. The method of claim 1 , further comprising:

building the voice action based on the recognition of the second portion of the voice input.

9. The method of claim 8 , wherein building the voice action is by a streaming semantic processor of the voice-enabled electronic device.

10. The method of claim 1 , wherein performance of the voice action comprises rendering, at the voice-enabled electronic device, of content that corresponds to the voice action.

11. A voice-enabled electronic device comprising:

one or more microphones;

memory storing instructions; and

one or more processors operable to execute the instructions stored in the memory, comprising instructions to:

generate, using a locally stored list of relevant context sensitive entities, voice to text model update data for a context sensitive parameter that is associated with the relevant context sensitive entities, wherein the voice to text model update data comprises decoding paths for the relevant context sensitive entities;

subsequent to generating the voice to text model update data:

receive a voice input via the one or more microphones, the voice input including an original request that includes first and second portions, the second portion including a first context sensitive entity among the relevant context sensitive entities; and

responsive to receiving the first portion of the voice input:

perform local processing of the first portion of the voice input;

determine during the local processing that the first portion is associated with the context sensitive parameter;

in response to determining that the first portion is associated with the context sensitive parameter:

 dynamically update a local voice to text model, used by the voice-enabled electronic device, using the locally generated voice to text model update data, wherein dynamically updating the local voice to text model facilitates recognition of the first context sensitive entity in performing local processing of the second portion of the voice input;

generate, utilizing the dynamically updated local voice to text model, a recognition of the second portion of the voice input; and

cause performance of a voice action that is based on the recognition of the second portion of the voice input.

12. The voice-enabled electronic device of claim 11 , wherein the instructions to perform the local processing include instructions to:

convert a digital audio signal of the voice input to text using a streaming voice to text module of the voice-enabled electronic device, wherein the streaming voice to text module dynamically generates a plurality of text tokens from the digital audio signal.

13. The voice-enabled electronic device of claim 11 , wherein the instructions to dynamically update the local voice to text model comprise instructions to augment the local voice to text model using the decoding paths.

14. The voice-enabled electronic device of claim 11 , wherein the instructions to determine during the local processing that the first portion is associated with the context sensitive parameter include instructions to determine that the first portion includes a request to play a media item, wherein the context sensitive parameter includes a media item identifier for use in identifying the media item, and wherein the plurality of context sensitive entities identify a plurality of media items playable by the voice-enabled electronic device.

15. The voice-enabled electronic device of claim 11 , wherein the instructions to determine during the local processing that the first portion is associated with the context sensitive parameter include instructions to determine that the first portion includes a request to communicate with a contact, wherein the context sensitive parameter includes a contact identifier for use in initiating a communication with the contact, and wherein the plurality of context sensitive entities identify a plurality of contacts accessible by the voice-enabled electronic device.

16. The voice-enabled electronic device of claim 11 , wherein the context sensitive parameter is a location-dependent parameter, and wherein the plurality of context sensitive entities identify a plurality of points of interest disposed in proximity to a predetermined location.

17. The voice-enabled electronic device of claim 16 , wherein the predetermined location comprises a current location of the voice-enabled electronic device, and the instructions further comprise instructions to:

in response to a determination that the first portion is associated with the context sensitive parameter, communicate the current location to an online service and prefetch from the online service voice to text model update data associated with the plurality of context sensitive entities.

18. The voice-enabled electronic device of claim 11 , wherein the instructions further comprise instructions to:

build the voice action based on the recognition of the second portion of the voice input.

19. The voice-enabled electronic device of claim 11 , wherein performance of the voice action comprises rendering, at the voice-enabled electronic device, of content that corresponds to the voice action.

20. A non-transitory computer readable storage medium storing computer instructions executable by one or more processors to perform a method comprising:

generating, using a locally stored list of relevant context sensitive entities, voice to text model update data for a context sensitive parameter that is associated with the relevant context sensitive entities, wherein the voice to text model update data comprises decoding paths for the relevant context sensitive entities;

subsequent to generating the voice to text model update data:

receiving a voice input with a voice-enabled electronic device, the voice input including an original request that includes first and second portions, the second portion including a first context sensitive entity among the relevant context sensitive entities; and

in the voice-enabled electronic device, and responsive to receiving the first portion of the voice input:

performing local processing of the first portion of the voice input;

determining during the local processing that the first portion is associated with the context sensitive parameter;

in response to determining that the first portion is associated with the context sensitive parameter:

dynamically updating a local voice to text model, used by the voice-enabled electronic device, using the locally generated voice to text model update data, wherein dynamically updating the local voice to text model facilitates recognition of the first context sensitive entity in performing local processing of the second portion of the voice input;

generating, utilizing the dynamically updated local voice to text model, a recognition of the second portion of the voice input; and

causing performance of a voice action that is based on the recognition of the second portion of the voice input.