Voice-controlled management of user profiles

1. A method of managing user profiles, comprising:

calculating, for each speaker model of a plurality of speaker models, a confidence measure representing a probability that the speaker model represents a speaker of a cluster of audio segments; and

responsive to the confidence measures of two or more of the speaker models representing probabilities that are each higher than a target probability, updating each of a plurality of user profiles based on a user preference assigned to the cluster of audio segments, each updated user profile being associated with a respective one of the two or more speaker models.

2. The method of claim 1 , further comprising performing speaker diarization on an audio stream to form at least one cluster of audio segments, wherein each cluster of audio segments comprises speech of a single speaker.

3. The method of claim 2 , wherein the performing speaker diarization comprises:

detecting speech active segments from the audio stream;

detecting speaker change points in the speech active segments to form audio segments of a single speaker; and

clustering audio segments of a same single speaker to form the at least one cluster of audio segments.

4. The method of claim 1 , further comprising updating the two or more speaker models based on the cluster of audio segments in response to the confidence measures calculated for the two or more speaker models each representing a probability that is higher than the target probability.

5. The method of claim 1 , further comprising assigning the user preference to the cluster of audio segments.

6. The method of claim 5 , wherein the assigning the user preference comprises:

performing automatic speech recognition on the cluster of audio segments to identify the user preference; and

assigning the identified user preference to the cluster of audio segments.

7. The method of claim 5 , wherein the user preference is selected from a group consisting of: a genre of a media, a name of a singer or band, a name of an actor or actress, a name of a director, a name of a sport team or athlete, a user equipment setting, and a web site address.

8. The method of claim 1 , further comprising identifying a speaker model of the two or more speaker models for which the calculated confidence measure represents a highest probability among the calculated confidence measures.

9. The method of claim 1 , further comprising:

calculating, for each of the plurality of speaker models, a further confidence measure representing a further probability that the speaker model represents a speaker of a further cluster of audio segments;

responsive to each of the further confidence measures representing a respective further probability that is not higher than the target probability:

creating a new speaker model;

updating a default user profile based on a user preference associated with the further cluster of audio segments; and

associating the updated default user profile with the new speaker model.

10. A device for managing user profiles, comprising:

processing circuitry;

memory containing instructions executable by the processing circuitry whereby the device is operative to:

calculate, for each speaker model of a plurality of speaker models, a confidence measure representing a probability that the speaker model represents a speaker of a cluster of audio segments; and

responsive to the confidence measures of two or more of the speaker models representing probabilities that are each higher than a target probability, update each of a plurality of user profiles based on a user preference assigned to the cluster of audio segments, each updated user profile being associated with a respective one of the two or more speaker models.

11. The device of claim 10 , wherein the instructions are such that the device is operative to perform speaker diarization on an audio stream to form at least one cluster of audio segments, wherein each cluster of audio segments comprises speech of a single speaker.

12. The device of claim 11 , wherein the instructions are such that the device is operative to:

detect speech active segments from the audio stream;

detect speaker change points in the speech active segments to form audio segments of a single speaker; and

cluster audio segments of a same single speaker to form the at least one cluster of audio segments.

13. The device of claim 10 , wherein the instructions are such that the device is operative to update the two or more speaker models based on the cluster of audio segments in response to the confidence measures calculated for the two or more speaker models each representing a probability that is higher than the target probability.

14. The device of claim 10 , wherein the instructions are such that the device is operative to assign the user preference to the cluster of audio segments.

15. The device of claim 14 , wherein the instructions are such that the device is operative to:

perform automatic speech recognition on the cluster of audio segments to identify the user preference; and

assign the identified user preference to the cluster of audio segments.

16. The device of claim 10 , wherein the instructions are such that the device is further operative to identify a speaker model of the two or more speaker models for which the calculated confidence measure represents a highest probability among the calculated confidence measures.

17. The device of claim 10 , wherein the instructions are such that the device is operative to:

calculate, for each of the plurality of speaker models, a further confidence measure representing a further probability that the speaker model represents a speaker of a further cluster of audio segments;

responsive to each of the further confidence measures representing a respective further probability that is not higher than the target probability:

create a new speaker model;

update a default user profile based on a user preference associated with the further cluster of audio segments; and

associate the updated default user profile with the new speaker model.