SYSTEM AND METHOD FOR LOW-LATENCY WEB-BASED TEXT-TO-SPEECH WITHOUT PLUGINS

• US 20130144624A1
• Filed: 12/01/2011
• Published: 06/06/2013
• Est. Priority Date: 12/01/2011
• Status: Active Grant

• Current Assignee: Cerence Operating Company (Cerence Inc.)
• Original Assignee: AT&T Intellectual Property I LP (AT&T, Inc.)
• Inventors: CONKIE, Alistair D., Mishra, Taniya, Charles Beutnagel, Mark
• Granted Patent: US 9,240,180 B2
• Time in Patent Office: Days
• US Class Current: 704/260
• CPC Class Codes: G10L 13/04 Details of speech synthesis systems, e.g. synthesiser structure or memory management
• CPC Class Codes: G10L 13/10 Prosody rules derived from text; Stress or intonation

Abstract

Disclosed herein are systems, methods, and non-transitory computer-readable storage media for reducing latency in web-browsing TTS systems without the use of a plug-in or Flash® module. A system configured according to the disclosed methods allows the browser to send prosodically meaningful sections of text to a web server. A TTS server then converts intonational phrases of the text into audio and responds to the browser with the audio file. The system saves the audio file in a cache, with the file indexed by a unique identifier. As the system continues converting text into speech, when identical text appears the system uses the cached audio corresponding to the identical text without the need for re-synthesis via the TTS server.

Claims

1. A method comprising:

receiving, from a client, text associated with a request for text-to-speech synthesis;

identifying a plurality of intonational phrases in the text;

generating a file containing text-to-speech data for a first intonational phrase of the plurality of intonational phrases, wherein the first intonational phrase is indexed by a unique identifier;

transmitting the file to the client in response to the request; and

generating files containing additional text-to-speech data for remaining intonational phrases of the plurality of intonational phrases, wherein each of the files is indexed by the unique identifier plus a respective offset.

2. The method of claim 1, wherein an intonational phrase is a phrase in which intonation within the phrase only depends on text inside the phrase.

3. The method of claim 1, wherein the file is indexed by a unique identifier.

4. The method of claim 1, wherein the file contains notification information.

5. The method of claim 1, wherein the unique identifier comprises a text identifier and an offset index.

6. The method of claim 1, wherein the additional file contains additional notification information.

7. The method of claim 1, wherein generating the additional file occurs while the web browser plays the text-to-speech data in the file.

8. The method of claim 1, wherein the file and the second file are stored in a cache.

9. The method of claim 1, further comprising transmitting one of the files to the web browser in response to an additional request.

10. The method of claim 1, wherein the notification information comprises synchronization data.

11. The method of claim 1, wherein boundaries between intonational phrases comprise silence.

12. The method of claim 1, further comprising:

receiving text-to-speech settings from the client; and

generating the file and the files based on the text-to-speech settings.

13. The method of claim 1, further comprising:

generating parallel versions of the file and the files using different text-to-speech voices.

14. A system comprising:

a processor;

a network interface;

a memory having stored therein instructions for controlling the processor to perform steps comprising;

receiving text from a user;

transmitting the text to a server as part of a request for text-to-speech synthesis;

receiving, from the server, a file containing a first intonational phrase indexed by a unique identifier;

playing the first intonational phrase; and

fetching an additional file for playback, wherein the additional file contains an additional intonational phrase indexed by the unique identifier plus an offset.

15. The system of claim 14, wherein the instructions are associated with a web browser.

16. The system of claim 15, wherein playing the first intonational phrase does not rely on a browser plugin.

17. The system of claim 14, fetching the additional file is based on client-side scripting.

18. The system of claim 14, further comprising:

receiving user input navigating to a different position within the text;

identifying a new offset for the different position; and

fetching a corresponding file from the server for playback based on the unique identifier and the new offset.

19. A non-transitory computer-readable storage medium having stored therein instructions which, when executed by a computing device, cause the computing device to perform steps comprising:

receiving, from a client, text associated with a request for text-to-speech synthesis;

identifying a plurality of intonational phrases in the text;

generating a file containing text-to-speech data for a first intonational phrase of the plurality of intonational phrases, wherein the first intonational phrase is indexed by a unique identifier;

transmitting the file to the client in response to the request; and

generating files containing additional text-to-speech data for remaining intonational phrases of the plurality of intonational phrases, wherein each of the files is indexed by the unique identifier plus a respective offset.

20. The non-transitory computer-readable storage medium of claim 19, the instructions further causing the computing device to perform steps comprising:

generating parallel versions of the file and the files using different text-to-speech voices.