Speech morphing communication system

• US 9,053,094 B2
• Filed: 10/30/2011
• Issued: 06/09/2015
• Est. Priority Date: 10/31/2010
• Status: Active Grant

• Current Assignee: Speech Morphing, Inc.
• Original Assignee: Speech Morphing, Inc.
• Inventors: Yassa, Fathy
• Primary Examiner(s): COLUCCI, MICHAEL C
• Application Number:
  US13/284,966
• Publication Number:
  US 20120109648A1
• Time in Patent Office: 1,318 Days
• Field of Search: 704/2, 704/9, 704/4, 704/277, 704/275, 704/270, 704/268, 704/266, 704/260, 704/258, 704/254, 704/250, 704/240, 704/235, 704/224, 704/221, 704/211, 704/209, 704/207, 704/10, 709/224, 705/17, 455/557
• US Class Current: 1/1
• CPC Class Codes: G06F 40/51 Translation evaluation
• CPC Class Codes: G06F 40/58 Use of machine translation, e.g. for multi-lingual retrieval, for server-side translation for client devices or for real-time translation
• CPC Class Codes: G10L 13/00 Speech synthesis; Text to speech systems
• CPC Class Codes: G10L 15/26 Speech to text systems G10L15/08 takes precedence

Abstract

A communication system is described. The communication system including an automatic speech recognizer configured to receive a speech signal and to convert the speech signal into a text sequence. The communication also including a speech analyzer configured to receive the speech signal. The speech analyzer configured to extract paralinguistic characteristics from the speech signal. In addition, the communication system includes a speech output device coupled with the automatic speech recognizer and the speech analyzer. The speech output device configured to convert the text sequence into an output speech signal based on the extracted paralinguistic characteristics.

Claims

1. A communication system comprising:

an automatic speech recognizer configured to receive a first speech signal of a first speaker and to convert said first speech signal into a text sequence;

a speech analyzer configured to receive said first speech signal, said speech analyzer configured to extract paralinguistic characteristics from said first speech signal; and

a speech output device coupled with said automatic speech recognizer and said speech analyzer, said speech output device configured to receive a second speech signal of a second speaker, convert said text sequence of said first speech signal into a converted speech signal of said second speaker based on one or more phonemes of said second speech signal, and convert said converted speech signal into an output speech signal of said second speaker by applying said extracted paralinguistic characteristics to said converted speech signal.

2. The communication system of claim 1, wherein said speech output device includes a personifier.

3. The communication system of claim 1, wherein said automatic speech recognizer converts said speech signal into said text sequence based on a first algorithm.

4. The communication system of claim 3, further comprising a second automatic speech recognizer, said second automatic speech recognizer configured to convert said first speech signal to a second text sequence based on a second algorithm.

5. The communication system of claim 4, further comprising a text sequence comparator configured to compare said text sequence with said second text sequence.

6. The communication system of claim 5, wherein said text sequence comparator generates an error corrected text sequence based on the comparison of said text sequence and said second text sequence.

7. The communication system of claim 5, wherein said text sequence comparator generates one or more confidence scores for said text sequence and for said second text sequence.

8. The communication system of claim 1, wherein said output device converts said text sequence into the converted speech signal using a text-to-speech algorithm.

9. A communication system comprising:

a memory;

one or more processors; and

one or more modules stored in memory and configured for execution by the one or more processors, the one or more modules comprising;

a speech input module configured to receive a first speech signal of a first speaker;

an automatic speech recognizer module coupled with said speech input module, said automatic speech recognizer module configured to convert said first speech signal into a text sequence;

a speech analyzer module coupled with said speech input module, said speech analyzer module configured to extract paralinguistic characteristics from said first speech signal; and

an output speech module configured to receive a second speech signal of a second speaker, convert said text sequence of said first speech signal into a converted speech signal of said second speaker based on one or more phonemes of said second speech signal, and convert said converted speech signal into an output speech of said second speaker by applying said paralinguistic characteristics to said converted speech signal.

10. The communication system of claim 9, wherein said output speech module includes a personifier.

11. The communication system of claim 9, wherein said automatic speech recognizer converts said speech signal into said text sequence based on a first algorithm.

12. The communication system of claim 11, further comprising a second automatic speech recognizer module, said second automatic speech recognizer module configured to convert first said speech signal to a second text sequence based on a second algorithm.

13. The communication system of claim 12, further comprising a text sequence comparator module configured to compare said text sequence with said second text sequence.

14. The communication system of claim 13, wherein said text sequence comparator module generates an error corrected text sequence based on the comparison of said text sequence and said second text sequence.

15. The communication system of claim 13, wherein text sequence comparator module generates one or more confidence scores for said text sequence and for said second text sequence.

16. The communication system of claim 9, wherein said output speech module converts said text sequence into the converted speech signal using a text-to-speech algorithm.

17. A speech morphing communication system comprising:

a speech input device configured to receive a first speech signal of a first speaker;

a first automatic speech recognizer and a second auk speech recognizer coupled with said input device, said first automatic speech recognizer configured to convert said first speech signal into a first text sequence and said second automatic speech recognizer configured to convert said first speech signal into a second text sequence;

a text sequence comparator coupled with said first automatic speech recognizer and said second automatic speech recognizer, said text sequence comparator configured to select one text sequence from among said first text sequence and said second text sequence;

a speech analyzer coupled with said speech input device, said speech analyzer configured to extract paralinguistic characteristics from said first speech signal; and

a speech output device configured to receive a second speech signal of a second speaker, convert said selected text sequence of said first speech signal into a converted speech signal of said second speaker based on one or more phonemes of said second speech signal, and convert said converted speech signal into an output speech signal of said second speaker by applying said paralinguistic characteristics to said converted speech signal.

18. A method for converting speech into text comprising:

receiving a first speech signal of a first speaker;

receiving a second speech signal of a second speaker;

extracting paralinguistic characteristics from said first speech signal;

converting said first speech signal to a text sequence;

transforming said text sequence of said first speech signal into a converted speech signal of said second speaker based on one or more phonemes of said second speech signal; and

converting said converted speech signal into an output speech signal of said second speaker by applying said extracted paralinguistic characteristics to said converted speech signal.

19. The method of claim 18 wherein, transforming includes using one or more text-to-speech algorithms.