MXPA97002228A - Instant cut of speech processing characteristics on a telep call - Google Patents

Abstract

The present invention relates to a speech processor that uses at least two speech processing features to improve the quality of speech signals received by a user during a telephone call. The speech processing characteristics are applied to speech signals. However, the user only hears speech signals affected by a speech processing feature until both characteristics have fully converged or ramp-up, and the two characteristics no longer interfere with each other. At this point, an "instant cut" of the second speech processing feature is activated. The instantaneous cut instantly switches to speech signals affected by both characteristics. This rapid transition makes speech processing features more noticeable to the user, and the user is not stuck to a period where features interfere. In addition, an optional audio indicator is generated before increasing the instantaneous cut, so that the user is alert to the instant cut and the characteristics of speech precessing are even more noticeable.

Description

INSTANT CUT OF SPEECH PROCESSING FEATURES ON A TELEPHONE CALL BACKGROUND OF THE INVENTION The present invention relates to improving the speech quality in a telephone call and more particularly to a method and apparatus that provide an instantaneous cut of speech processing features in a telephone call. It is well known in the telecommunications art to apply speech processing features in a telephony network in order to improve the quality of speech signals. Some features provide virtually all of their intended effect immediately upon activation. These characteristics are referred to as "non-adaptive" and include, for example, pre-emphasis compensating filters. Other features, however, gradually and uniformly apply their effect, that is, they "rise in ramp" after activation.

These characteristics are referred to as "adaptive" and include for example automatic gain control, background interference compensation, interference reduction and echo cancellation. REF: 24290 It is known that more than one speech processing feature can be applied in a telephony network. For example, the US patent. No. 5,195,132 issued to Bowker et al. On March 16, 1993, describes the use of both echo cancellation and digital filtering to improve the quality of the speech signal. However, a problem that to date has not been recognized in the telecommunications technique arises when more than one speech processing feature is applied to a telephony network, especially with telephony networks that use echo cancellers. This problem can be seen in Figure 1 which shows a graph of a particular telephone call starting at time t0. Curve 8 represents echo cancellation in the network. As is known in the art, echo cancellation requires time after the start of a call to converge or completely "ascend in ramp", and in Figure 1, or in the convergence of curve 8 occurs at time tj .. curve 9 represents another adaptive process such as background interference compensation that consumes time ta to ramp up. A problem arises through the length of time ti-to when the ramp up of both processes translates. During this period, the processes interfere with each other and the quality of calls is severely degraded. Therefore, there is a need in the art to provide multiple speech processing features to a telephone network without initially degrading the quality of the call. Another problem with the techniques described in the art for applying speech procedure features to a telephony network involves the user's perception of the effect of these characteristics. The beginning of telecommunications, the speech processing features have always been provided at the beginning of the call and the motivation of the designers of telecommunications systems has always been to reduce the ramp-up time of the characteristics, in such a way that the transition The integral effectiveness of the features is minimal, perceptible by the client. For example in the U.S. Patent. No 5,001,701 awarded to Gay on March 19, 1991, describes the use of real-time allocation between sub-bands to achieve faster total convergence of echo cancellation. However, we have found that if the speech processing features are provided directly from the beginning of the call, with fast ramp-up times, users may not attribute the higher quality call to the presence of speech processing features . Therefore, there is a need to alert the user that the speech processing features that improve the speech signal quality are applied to a particular call.

COMPENDI EP The Invention According to one embodiment of the present invention, two speech processing features are applied to the speech signals of a telephone call. However, the user only hears speech signals affected by a speech processing feature until features have fully converted or ramp-up, and the two features no longer interfere with each other. At that point, an "instant cut" of the second speech processing feature is activated. The instantaneous cut instantly switches to speech signals affected by both characteristics. This rapid transition makes the speech processing features more noticeable to the user and the user is not subject to the period in which the features interfere. In another embodiment of the present invention, two speech processing features are applied to the speech signals of a telephone call. However, the user hears speech signals unaffected by either speech processing feature until both features have fully converged or ramp-up, and the two features no longer interfere with each other. At that point, an "instant cut" of both speech processing features is activated.

In another embodiment of the present invention, an audio injector is generated before the instantaneous cut is implemented, such that the user is alert to the instantaneous cutoff and the characteristics of the speech processing are even more noticeable. The above-described features of the present invention are not found in the prior art because of the conventional wisdom in the telecommunications art is to minimize as much as possible the tendency to intrusion and noticeable to the user of speech processing characteristics. In contrast, in the present invention, the instantaneous cut of the audio indicator increases the intrusive and remarkable speech processing characteristics. BRIEF DESCRIPTION OF THE DRAWINGS Figure 1 is a graph illustrating two speech processing characteristics that are transposed. Figure 2 is a block diagram of a speech processor embodiment of the present invention. Figure 3 is a block diagram of another embodiment of the speech processor of the present invention. Figure 4 is a block diagram of another embodiment of the speech processor of the present invention. DESCRIPTION OF THE INVENTION For clarity of explanation, the illustrative embodiment of the present invention is presented comprising individual functional blocks (including functional blocks labeled "processors"). The functions that these blocks represent can be provided through the use of either shared or dedicated physical equipment, including but not limited to, physical equipment capable of executing software. For example, the functions of the processors presented in Figure 2 can be provided by a simple shared processor. (Use of the term "processor" shall not be considered to refer exclusively to physical equipment capable of running software). Exemplary embodiments may comprise digital signal processor (DSP) hardware such as DSP16 or DSP32C from Lucent Technologies, read-only memory (ROM) for storing software performing the operations discussed below, and random access memory (RAM) for store the DSP results. Modalities of physical equipment with very large scale integration (VLSI) as well as VLSI circuits tailored in combination with a general purpose DSP circuit can also be provided. With reference in detail to the drawings, where parts. similar are designated by like reference numbers, a block diagram of a speech processor 15 according to an embodiment of the present invention is illustrated in Figure 2. In Figure 2, "income talk" refers to the speech signal before processing while "output speech" refers to the speech signal after processing. The speech processor 15 includes an echo canceller 10 that performs echo cancellation in speech input. The echo canceller input 10 is coupled to the ingress speech path and the output is coupled to the input of both the fixed delay unit 18 and a speech enhancement processor 20. The speech improvement processor 20 implements one or more speech processing algorithms to process income speech. In one embodiment, the speech enhancement processor 20 performs background interference compensation in the ingress speech. The fixed delay unit 18 portrays the speech path by an amount equal to the total delay introduced by the speech enhancement processor 20. The output of the fixed delay unit 18 and the speech enhancement processor 20 are selectively coupled through from a switch 22 to the output speech path. The speech processor 15 further includes a delay synchronizer 14. The delay synchronizer 14 is coupled to the switch 22 and includes a re-initialization feed 16. The delay synchronizer 14 can already configure the switch 22 such that the unit of fixed delay 18 is coupled to the output speech path (the "first position") or in such a way that the speech enhancement processor 20 is coupled to the output speech path (the "second position"). When a re-initialization signal is received by the re-initialization feed 16, the delay synchronizer 14 waits for a fixed period of time and then sets the switch 22 to the second position. A telephone call is initiated for the purposes of the speech processor 15, after the calling party has completed the dialing. The switch 22 initially configures in the first position before the call is initiated. Therefore, initially the output speech signals will only be affected by echo cancellation (and a delay). A re-initialization signal is either sent to the re-initialization feed 16 when a call is initiated or when the called party has answered the call. When the delay synchronizer 14 rotates, the switch 22 is switched, or "instantaneously cut" to the second position and the output speech signals are then affected by both echo cancellation and background interference compensation. The amount of time that the delay synchronizer 14 waits until it expires, is adjusted such that the echo cancellation has fully converged and the background interference compensation has been fully ramp-up. In one embodiment, if the re-initialization signal is sent to the re-initialization feed 16 when the call is initiated, the delay synchronizer 14 is set to expire in approximately 55 seconds.; if the re-initialization signal is sent to the re-initialization feed 16 when the called party has answered the call, the delay synchronizer 14 is set to expire in approximately 7 seconds. The result is that the quality of the speech signals received by the user increases suddenly when the delay synchronizer 14 expires and the signals are affected by the completely high background interference compensation in ramp. In addition, the user is not subjected to degraded speech signals during the period in which the two speech processing characteristics are transposed, i.e. during time t, a to in Figure 1. Figure 3 is a block diagram of a speech processor 32 according to another embodiment of the present invention. The speech processor 32 is identical to the speech processor 15 illustrated in Figure 2 except that the speech processor 32 includes an audio logo generator 30 coupled to the delay synchronizer 14 and the speech speech output. The audio logo generator 30, when triggered by the expiration of the delay synchronizer 14, generates an audio logo and adds it to the output speech. The audio logo alerts the customer that the phone call is being cut instantly and the speech signals are now affected by both echo cancellation and background interference compensation. Therefore, the audio logo causes the background interference compensation effect to be even more noticeable to the user. Each component of the present invention has been illustrated in block diagram form to facilitate clarity of the invention. The functionality of each component can be implemented by conventional equipment that is known to persons of ordinary skill in the art. Furthermore, what has been described is merely illustrative of the application of the principles of the present invention. Other assemblies and methods may be implemented by those skilled in the art, without departing from the spirit and scope of the present invention. For example, instead of the user initially receiving speech signals affected by echo cancellation, the user may initially receive speech signals not affected by any speech processing feature. The speech signals affected by both echo cancellation and background interference compensation and any other speech processing feature can all be instantly cut off on the speech signals immediately. Figure 4 is a block diagram illustrating one embodiment of this capability wherein the speech signal improvement that is provided by the echo canceller 10 and the speech improvement processor 20 are instantly cut by the switch 22, simultaneously low the synchronizer control 14.

It is noted that in relation to this date, the best method known to the applicant to carry out the aforementioned invention, is that which is clear from the present description of the invention. Having described the invention as above, property is claimed as contained in the following:

Claims (20)

1. CLAIMS 1.- A method for using a plurality of speech processing features to improve the quality of speech signals received by a user during a telephone call in a telephony network, wherein the network can be switched either to a mode not improved where the user receives speech signals not affected by the application of a second speech processing feature, or to an improved mode in which the user receives speech signals affected by the application of the second speech processing feature, characterized in that it comprises the steps of: switching the network to unimproved mode; initiating the application of the first speech processing feature to the speech signals of the telephone call and initiating the application of the second speech processing feature to the speech signals of the telephone calls; and switching the network to the enhanced mode at the end of a first duration of time after the start of the application of the second speech processing feature.
2. 2. A method according to claim 1, characterized in that the first speech processing feature interferes with the second speech processing feature for a second duration of time after the telephone call is initiated, and wherein the The first duration of time is greater than the second duration of time.
3. 3. A method according to claim 2, characterized in that it further comprises the step of: delaying the time that the speech signals are received by the user during the first duration of time.
4. 4. A method according to claim 3, characterized in that it also comprises the step of: sending an audio alert to the end user of the first time duration.
5. 5. A method according to claim 1, characterized in that in the unimproved mode, the user receives speech signals not affected by the application of the first speech processing feature, and in the enhanced mode, the user receives signals speech affected by the application of the first speech processing feature.
6. 6. A method according to claim 1, characterized in that in the unimproved mode, the user receives speech signals affected by the application of the first speech processing feature, and in the enhanced mode, the user receives signals of speech. was affected by the application of the first speech processing feature.
7. 7. - A method according to claim 1, characterized in that the first speech processing feature is echo cancellation and the second speech processing feature is background interference compensation.
8. 8. A method for using a plurality of speech processing features to improve the quality of speech signals received by a user during a telephone call in a telephony network, wherein the network can be switched to either a non-dialing mode. improved, wherein the user receives speech signals affected by the application of a second speech processing feature, or an improved mode where the user receives speech signals affected by the application of the second speech processing feature, and in wherein the network is in the unimproved mode when the call is initiated, characterized in that it comprises the steps of: applying a first speech processing feature to the speech signals when the telephone call is initiated; apply the second speech processing feature to the speech signals when the telephone call is initiated; and switching the network to the enhanced mode at the end of a first duration of time after the telephone call is initiated.
9. 9. - A method according to claim 8, characterized in that the first speech processing feature interferes with the second speech processing feature with a second duration of time after the telephone call is initiated, wherein the first duration of time is greater than the second duration of time.
10. 10. A method according to the claim 9, characterized in that it further comprises the step of: delaying the time that the speech signals are received by the user during the first duration of time.
11. 11.- A method in accordance with the claim 10, characterized in that it further comprises the step of: sending an audio signal to the user at the end of the first time duration.
12. 12. A method according to claim 8, characterized in that in the unimproved mode, the user receives speech signals not affected by the application of the first speech processing feature, and in the enhanced mode, the user receives signals speech affected by the application of the first speech processing feature.
13. 13. A method according to claim 8, characterized in that in the unimproved mode, the user receives speech signals affected by the application of the first speech processing feature, and in the enhanced mode, the user receives signals of speech. was affected by the application of the first speech processing feature.
14. 14.- A method according to the claim 8, characterized in that the first speech processing feature is echo cancellation and the second speech processing feature is background interference compensation.
15. 15. Speech processor for improving the quality of speech signals received by a user during a telephone call, characterized in that it comprises: a first processor for speech improvement that applies a first speech processing feature to the speech signals; a second processor for speech improvement that applies a second speech processing feature to the speech signals; and a switch switches the speech processor from an enhanced mode where the user receives speech signals not affected by the application of the second speech processing feature, to an improved mode where the user receives speech signals affected by the application of the second speech processing feature.
16. 16. - The speech processor according to claim 15, characterized in that it further comprises: a fixed delay unit that delays the time in which the speech signals are received by the user.
17. 17. The speech processor according to claim 16, characterized in that it further comprises: an audio indicator generator that sends an audio alert to the user.
18. 18. The speech processor according to claim 15, characterized in that in the unimproved mode, the user receives speech signals not affected by the application of the first speech processing feature, and in the enhanced mode, the user receives speech signals affected by the application of the first speech processing feature.
19. 19. The speech processor according to claim 15, characterized in that in the unimproved mode, the user receives speech signals affected by the application of the first speech processing feature, and in the enhanced mode, the user receives speech signals affected by the application of the first speech processing feature.
20. 20. A method for applying a plurality of speech processing features to improve the quality of a speech signal communicated over a communication network, wherein the speech signal is in an improved form by virtue of the application of both as second features of speech processing, the method is characterized in that it comprises: providing the speech signal to a party in an unimproved form; and waiting a period of time after the speech signal is provided to the party in the unimproved form and after the period, automatically providing the speech signal to the party in an improved manner.