Silence editing speech processor

• US 4,449,190 A
• Filed: 01/27/1982
• Issued: 05/15/1984
• Est. Priority Date: 01/27/1982
• Status: Expired due to Term

• Current Assignee: Bell Telephone Laboratories, Inc. (Nokia Corporation)
• Original Assignee: Bell Telephone Laboratories, Inc. (Nokia Corporation)
• Inventors: Flanagan, James L., Johnston, James D.
• Primary Examiner(s): KEMENY, EMANUEL
• Application Number:
  US06/343,238
• Time in Patent Office: 839 Days
• Field of Search: 364/513.5, 364/200 MS File, 364/900 MS File, 381/29-35, 381/46, 381/106, 375/122
• US Class Current: 706/22
• CPC Class Codes: G10L 25/78 Detection of presence or absence of voice signals switching of direction of transmission by voice frequency in two-way loud-speaking telephone systems H04M9/10
• CPC Class Codes: H04B 14/068 with adaptive feedback

Abstract

In an ADPCM system, improved detection of silence intervals in a speech signal is attained by detecting the level of the logarithm step-size signal (d_n), which is representative of the energy of the speech samples. A speech pattern is converted into a sequence of adaptive digital codes. Intervals of silence in the pattern are detected and a digital code representative of each silence interval is generated. The adaptive digital codes and the silence interval codes are combined to form a digitally coded signal representative of the pattern. The conversion of the pattern to adaptive digital codes includes forming a signal corresponding to the adaptation step-size for each digital code. The silence interval detection includes producing first and second threshold signals. A silence interval signal is initiated when the adaptation step-size corresponding signal diminishes below the first threshold and the silence interval is terminated when the adaptation step-size corresponding signal increases above the second threshold after the silence interval initiation.

Claims

1. A speech processing system comprising:

means for converting a speech pattern into a sequence of adaptive digital coded signals;

means for detecting intervals of silence in said speech pattern;

means responsive to each detected silence interval for generating a digital coded signal representative thereof; and

means for combining said adaptive digital coded signals and said silence interval representative digital coded signals to form a digital signal corresponding to said pattern;

said converting means comprising;

means for forming a signal corresponding to the adaptation step-size for each adaptive digital coded signal;

means for generating first and second threshold signals, said second threshold signal being greater than said first threshold signal;

said silence interval detecting means comprising;

means responsive to said adaptation step-size corresponding signal diminishing below said first threshold signal for generating a signal indicative of a silence interval; and

means jointly responsive to said silence interval signal and said adaptation step-size corresponding signal increasing above said second threshold signal for terminating said silence interval signal.

2. A speech processing system according to claim 1 wherein said silence interval signal generating means comprises:

means responsive to said first threshold signal exceeding said step-size corresponding signal for producing a third signal;

means responsive to said third signal for initiating said silence interval signal;

means responsive to said step-size corresponding signal exceeding said second threshold signal for generating a fourth signal; and

means jointly responsive to said silence interval signal and said fourth signal for terminating said silence interval signal.

3. A speech processing system according to claim 2 wherein said means for forming said adaptation step-size corresponding signal comprises means for forming a signal representative of the logarithm of the step-size for each adaptive digitally coded signal.

4. A speech processing system according to claim 1, 2 or 3 wherein said threshold signal generating means comprises:

means for generating first and second predetermined level signals;

means responsive to the sequence of step-size corresponding signals for producing a signal representative of the maximum step-size corresponding signal in said sequence of step-size corresponding signals;

means jointly responsive to said maximum step-size corresponding signal and said predetermined first level signal for producing an adaptive first threshold level signal; and

means jointly responsive to said maximum step-size level corresponding signal and said second predetermined level signal for generating an adaptive second threshold level signal.

5. A method for processing speech comprising the steps of:

converting a speech pattern into a sequence of adaptive digital coded signals;

detecting intervals of silence in said speech pattern;

generating a digital coded signal representative of each detected silence interval; and

combining said adaptive digital coded signals and said silence interval representative coded signals to form a digital signal representative of the speech pattern;

said speech pattern converting step comprising;

forming a signal corresponding to the adaptation step-size for each adaptive digital coded signal;

generating first and second threshold signals;

said second threshold signal being greater than said first threshold signal; and

said silence interval detecting step comprising generating a signal indicative of said silence interval responsive to said adaptation step-size corresponding signal diminishing below said first threshold signal and terminating said silence interval signal responsive to said silence threshold signal and said adaptation step-size corresponding signal increasing above said second threshold signal.

6. A method for processing speech according to claim 5 wherein said silence interval signal generating step comprises producing a third signal responsive to said first threshold signal exceeding said step-size corresponding signal;

initiating said silence interval responsive to said third signal;

generating a fourth signal responsive to said step-size corresponding signal exceeding said second threshold signal; and

terminating said silence interval signal jointly responsive to said silence interval signal and said fourth signal.

7. A method for processing speech according to claim 6 wherein forming said adaptation step-size corresponding signal comprises forming a signal representative of the logarithm of the step-size for each adaptive digital coded signal.

8. A method for processing speech according to claims 5, 6 or 7 wherein said threshold signal generating step comprises:

generating first and second predetermined level signals;

producing a signal representative of the maximum step-size corresponding signal in said sequence of step-size corresponding signals;

producing an adaptive first threshold signal jointly responsive to said maximum step-size corresponding signal and said predetermined first level signal; and

producing an adaptive second level threshold signal responsive to said maximum step-size corresponding signal and said second predetermined level signal.

9. A voice storage system comprising:

means for converting a speech message into a sequence of adaptive digital coded signals;

means for detecting intervals of silence in said speech message;

means responsive to each detected silence interval for generating a coded signal representative thereof; and

means responsive to said adaptive digital coded signals and said silence interval coded signals for storing a digital coded signal sequence representative of said speech message;

said converting means comprises;

means for forming a signal corresponding to the adaptive step-size for each adaptive digital coded signal; and

said silence interval detecting means comprises means for generating first and second threshold signals, said second threshold signal being greater than said first threshold signal;

means responsive to said adaptive step-size corresponding signal being less than said first threshold signal for initiating a signal indicative of a silence interval; and

means responsive to said silence interval signal and said adaptive step-size signal exceeding said second threshold signal for terminating said silence interval signal.

10. A voice storage system according to claim 9 wherein said means for forming said adaptive step-size corresponding signal comprises means for generating a signal representative of the logarithm of said adaptive step-size signal.

11. A voice storage system according to claim 9 wherein said first and second threshold signal generating means includes:

means for generating first and second predetermined level signals;

means for selecting the maximum adaptive step-size corresponding signal in the sequence of adaptive step-size corresponding signals at the occurrence of each adaptive digital coded signal;

means responsive to the first predetermined level signal and said selected maximum adaptive step-size corresponding signal for generating an adaptive first threshold signal; and

means jointly responsive to said second predetermined level signal and said selected maximum adaptive step-size corresponding signal for generating an adaptive second threshold signal.

12. A voice storage system according to claim 9, claim 10 or claim 11 furhher comprising means responsive to said stored digital code sequence for constructing a speech pattern corresponding to said speech message.

13. A voice storage method comprising the steps of:

converting a speech message into a sequence of adaptive digital coded signals;

detecting intervals of silence in said speech message;

generating a coded signal representative of each detected silence interval;

storing a digital coded signal representative of the speech message responsive to said adaptive digital coded signals and said silence interval representative signals;

said converting step comprising forming a signal corresponding to the adaptive step-size for each adaptive digital coded signal; and

said silence interval detecting step comprising;

generating first and second threshold signals, said second threshold signal being greater than said first threshold signal;

initiating a signal indicative of a silence interval responsive to said adaptive step-size signal diminishing below said first threshold signals; and

terminating the silence interval signal responsive to said silence interval signal and the adaptive step-size signal exceeding said second threshold signal.

14. A voice storage method according to claim 13 wherein the forming of said adaptive step-size corresponding signal comprises forming a signal representative of the logarithm of the adaptive step-size signal.

15. A voice storage method according to claim 13 or claim 14 wherein said first and second threshold signal generating step comprises:

generating first and second predetermined level signals;

said second predetermined level signal being greater than said first predetermined level signal;

selecting the maximum adaptive step-size corresponding signal in the sequence of adaptive step-size corresponding signals at the occurrence of each adaptive digital coded signal;

generating a first adaptive threshold signal responsive to the selected maximum step-size signal and said first predetermined level signal; and

generating a second adaptive threshold signal response to said selected maximum adaptive step-size signal and said second predetermined level signal.