Method and system for training a text-to-speech synthesis system using a specific domain speech database
View Patent ↗A method and system are disclosed that train a text-to-speech synthesis system for use in speech synthesis. The method includes generating a speech database of audio files comprising domain-specific voices having various prosodies, and training a text-to-speech synthesis system using the speech database by selecting audio segments having a prosody based on at least one dialog state. The system includes a processor, a speech database of audio files, and modules for implementing the method.
1. A system for training a text-to-speech synthesis system, the system comprising:
a processor;
a speech database of audio files comprising domain-specific voices having various prosodies; and
a module configured to control the processor to train a text-to-speech synthesis system using the speech database by selecting audio segments having a prosody based on at least one dialog state.
2. The system of claim 1 , wherein the dialog state is one of a beginning, middle and end of a dialog.
3. The system of claim 1 , further comprising a module configured to control the processor to train the text-to-speech synthesis system using the speech database by selecting audio segments having a prosody based on at least one speech act.
4. The system of claim 3 , wherein the speech act is based on at least one of syntax and semantics of a user's input.
5. The system of claim 1 , wherein the module configured to control the processor to train a text-to-speech synthesis system selects audio segments based on at least one of discourse and semantics.
6. The system of claim 1 , further comprising:
a module configured to control the processor to update the speech database using dialogs conducted with user's using the text-to-speech system.
7. The system of claim 1 , further comprising a module configured to control the processor to select audio segments based on a user's profile.
8. The system of claim 1 , further comprising:
a module configured to control the processor to tag audio files based on prosody for storage in the speech database.
9. A non-transitory computer-readable storage medium storing instructions for controlling a computing device for use in training a text-to-speech speech synthesis, the instructions comprising:
generating a speech database of audio files comprising domain-specific voices having various prosodies; and
training a text-to-speech synthesis system using the speech database by selecting audio segments having a prosody based on at least one dialog state.
10. The non-transitory computer-readable storage medium of claim 9 , wherein the dialog state is one of a beginning, middle and end of a dialog.
11. The non-transitory computer-readable storage medium of claim 9 , further comprising training the text-to-speech synthesis system using the speech database by selecting audio segments having a prosody based on at least one speech act.
12. The non-transitory computer-readable storage medium of claim 11 , wherein the speech act is based on at least one of syntax and semantics of a user's input.
13. The non-transitory computer-readable storage medium of claim 9 , further comprising selecting audio segments based on at least one of discourse and semantics.
14. The non-transitory computer-readable storage medium of claim 9 , further comprising:
updating the speech database using dialogs conducted with user's using the text-to-speech system.
15. The non-transitory computer-readable storage medium of claim 9 , further comprising selecting audio segments based on a user's profile.
16. The non-transitory computer-readable storage medium of claim 9 , further comprising:
tagging audio files based on prosody for storage in the speech database.
17. A method for training a text-to-speech speech synthesis system, the instructions comprising:
generating, via a processor, a speech database of audio files comprising domain-specific voices having various prosodies; and
training, via a processor, a text-to-speech synthesis system using the speech database by selecting audio segments having a prosody based on at least one dialog state.
18. The method of claim 17 , wherein the dialog state is one of a beginning, middle and end of a dialog.
19. The method of claim 17 , further comprising training the text-to-speech synthesis system using the speech database by selecting audio segments having a prosody based on at least one speech act.
20. The method of claim 19 , wherein the speech act is based on at least one of syntax and semantics of a user's input.
21. The method of claim 17 , further comprising:
updating the speech database using dialogs conducted with user's using the text-to-speech system.
22. The method of claim 17 , further comprising:
tagging audio files based on prosody for storage in the speech database.