Speech system

• US 20030028380A1
• Filed: 08/02/2002
• Published: 02/06/2003
• Est. Priority Date: 02/02/2000
• Status: Abandoned Application

• Current Assignee: Famoice Technology Pty Limited
• Original Assignee: Famoice Technology Pty Limited
• Inventors: Freeland, Warwick Peter, Dixon, Ian Edward, Brien, Glenn Charles
• Application Number:
  US10/211,637
• Publication Number:
  US 20030028380A1
• Time in Patent Office: Days
• US Class Current: 704/260
• CPC Class Codes: G10L 13/00 Speech synthesis; Text to speech systems

Abstract

A system for generating an audio message over a communications network that is at least partly in a voice representative of a character generally recognizable to a user. Either a voice message or text based message may be used to construct the audio message. Specific recordings of well known characters is stored in a storage means and background sound effects can be inserted into the audio message which are stored in database. The audio message is constructed by any one of the processing means and transmitted to a recipient for play back on a processing terminal.

Claims

1. A method of generating an audio message, comprising the steps of:

providing a text-based message; and

generating said audio message based on said text-based message;

wherein said audio message is at least partly in a voice which is representative of a character generally recognisable to a user.

2. A method according to claim 1 wherein said character is selected from a predefined list of characters, each character in said list being generally recognisable to a user.

3. A method according to either claim 1 or claim 2 wherein said generating step uses a textual or encoded database which indexes speech units with corresponding audio recordings representing said speech units.

4. A method according to either claim 1 or claim 2 wherein said generating step comprises concatenating together one or more audio recordings of speech units, the sequence of the concatenated audio recordings being determined with reference to indexed speech units associated with one or more of the audio recordings in said sequence.

5. A method according to claim 3 further comprising the step of substituting words in said text-based message that do not have corresponding audio recordings of suitable speech units with substitute words that do have corresponding audio recordings.

6. A method according to claim 3, wherein said speech units represent any one or more of the following:

words, phones, sub-phones, multi-phone segments of speech.

7. A method according to claim 3 wherein said speech units cover the phonetic and prosodic range required to generate said audio message.

8. A method according to claim 5 wherein the substituted words are replaced with support words that each have suitable associated audio recordings.

9. A method according to claim 1 wherein after the step of providing said text-based message the method further comprising the step of converting said text-based message into a corresponding text-based message which is used as the basis for generating said audio message.

10. A method according to claim 9 wherein said step of converting said text-based message to a corresponding text-based message includes substituting said original text-base message with a corresponding text-based message which is an idiomatic representation of said original text-based message.

11. A method according to claim 10 wherein said corresponding text-based message is in an idiom which is attributable to, associated with or at least compatible with said character.

12. A method according to claim 10 wherein said corresponding text-based message is in an idiom which is intentinally incompatible with said character or attributable to or associated with a different which is generally recognisable by a user.

13. A method according to claim 1 wherein said audio message is generated in multiple voices, each voice representative of a different character which is generally recognisable to a user.

14. A method according to claim 1 wherein after the step of providing said text-based message the method further comprising the step of converting only a portion of said text-based message into a corresponding text-based message which is an idiomatic representation of the original text-based message.

15. A method according to claim 1 wherein said generating step includes randomly inserting particular vocal expressions or sound effects between certain predetermined audio recordings from which the audio message is composed.

16. A method according claim 1 wherein said text-based message is generated from an initial audio message from said user using voice recognition and subsequently used as the basis for generating said message in a voice representative of a generally recognisable character.

17. A method according to claim 1 further comprising the step of said user applying one or more audio effects to said audio message.

18. A method according to claim 17 wherein said one or more audio effects includes background sound effects to give the impression that the voice of the character emanates from a particular environment.

20. A computer program comprising computer program code to control a processing means to execute a procedure for generating an audio message according to the method of claim 1.

19. A method for generating an audio message which is at least partly in a voice representative of a character generally recognisable to a user, said method comprising the following steps:

transmitting a message request over a communications network;

processing said message request and constructing said audio message in at least partly a voice representative of a character generally recognisable to a user; and

forwarding the constructed audio message over said communication network to one or more recipients.