WO2015135748A1

WO2015135748A1 - Methods and devices for situation-adequate notifications

Info

Publication number: WO2015135748A1
Application number: PCT/EP2015/053838
Authority: WO
Inventors: Frank Dawidowsky; Michael Enenkl; Fritz Hohl; Thomas Kemp; Stefan Uhlich; Wilhelm Hagg
Original assignee: Sony Corporation; Sony Deutschland Gmbh
Priority date: 2014-03-11
Filing date: 2015-02-24
Publication date: 2015-09-17

Abstract

It is described an electronic device comprising a situation recognizer module configured to determine situation information from characteristics of a captured input signal, an adequateness decider module configured to determine an adequateness parameter from the situation information, and a notification output module configured to control the outputting of a notification based on the adequateness parameter.

Description

METHODS AND DEVICES FOR SITUATION-ADEQUATE

NOTIFICATIONS

The invention relates to methods and devices for providing situation-adequate notifications, for example speech notifications.

BACKGROUND

In some situations automatic speech output or other notification signals generated by devices such as navigation devices, mobile phones or TV sets can be disturbing. For example, an automatic speech device might interrupt other speakers in their flow of conversation. Also the output volume of the speech device might not be adequate. It might for example be unnatural if the speech device's output is significantly louder than the conversation of people.

Currently, a user of a mobile device has typically the manual choice of either let the device play out some fixed audio when some notification needs to be done (e.g. in case of a phone call, the reception of a text message, or an emerging event in the calendar), to let the mobile device vibrate, or to do nothing at all.

Furthermore, with current automatic speech devices, important speech messages might even not be recognized by humans, e.g. a navigation device user might miss a turn instruction if he is engaged in a communication.

In the light of this it is an object of the invention to provide improved notification devices and methods.

SUMMARY

Methods and devices for providing situation-adequate notifications are disclosed and defined in the appended set of claims.

According to a first aspect it is disclosed an electronic device comprising a situation recognizer module configured to determine situation information from characteristics of a captured input signal, an adequateness decider module configured to determine an adequateness parameter from the situation information, and a notification output module configured to control the outputting of a notification based on the adequateness parameter. According to a further aspect it is disclosed a method comprising: determining situation information from characteristics of a captured input signal, determining an adequateness parameter from the situation information, and controlling the outputting of a notification based on the adequateness parameter.

Further aspects are set forth in the dependent claims, the following description and the drawings.

BRIEF DESCRIPTION OF THE DRAWINGS

Details of the disclosed methods and devices will become more apparent from the following description of embodiments in connection with the accompanying drawings, of which

Fig. 1 schematically shows an embodiment of an electronic device for providing situation-adequate notifications;

Fig. 2 schematically shows an embodiment of a method for providing situation-adequate notifications;

Fig. 3 schematically shows an embodiment of a method for an automatic speech device trying to get the floor in order to output a speech notification; and

Fig. 4 schematically shows an embodiment of a method for an automatic speech device trying to get the floor in dependence of the urgency of a notification.

DETAILED DESCRIPTION OF THE EMBODIMENTS

Aspects of embodiments of methods and devices for providing situation- adequate notifications are now described.

A method for providing situation-adequate notifications comprises determining situation information from characteristics of a captured input signal, determining an adequateness parameter from the situation information, and controlling the outputting of a notification based on the adequateness parameter.

The method may be used by automatic speech devices, for example in a navigation device or a mobile communication device. Such speech devices may output notifications in form of computer generated speech. The method may also be used in devices which provide notifications in form of a sound signal, a vibration alert, a visual signal or other kind of notifications. Typically, the notifications address one or more user of a device. In a navigation device the addressed user is for example a car driver. By taking into account situation information, user notification can be done in a more natural way by adapting to the environment, and e.g. by considering "etiquette". When two people are talking to each other they adapt their conversation to each other in different ways. Such an adaptation can for example be implemented by electronic devices when doing speech output. For example, an automatic speech device may take into account the situation it is used in order to determine an adequateness parameter which describes how adequate it is to output a notification at a specific point in time, under specific conditions, etc.

Controlling the outputting of a notification may be based on the adequateness parameter and the situation information. For example, the output of a speech device may not be the same, whether the device is used during a meeting, in a car, or outside.

A natural communication typically requires from a speaker to get the floor before starting to talk. This means people usually wait until the others finished talking. A device such as a navigation device might act in a similar way.

If there is no break for some time because people do not stop talking, the method may use different mechanisms in order to get the floor. People try to interrupt politely by issuing filler words such as "urn", "well", by calling the first name of the person that they want to talk to or they harrumph to indicate that they try to get the floor. Another less polite way is to start talking and repeating the beginning of the utterance until the other person becomes silent. This might sound like "I"... "I want"... "I want" ... "I want to say that I do not agree with your statement". Finally people increase the volume of their attempt to interrupt when they don't succeed in getting the floor.

A device such as a navigation device might act in a similar way. If people continue to talk, the device might slowly get the attention by issuing a phrase such as "Please" and then wait for people interrupting their flow of speech. If this is not successful the attempt might be repeated with an increased volume.

In order to see whether this active approach to get the floor was successful, feedback as that the person is looking to the device (detected by an eye tracker) or says something like "yes", "what's up" can be used.

The method for providing situation-adequate notifications may further comprise outputting a notification signal which has the intention to interrupt the user.

Determining situation information may comprise at least one of detecting a speech pause in the captured input signal, detecting a predefined speech sequence in the captured input signal, detecting anything that indicates from a human communication politeness aspect that it is currently not so adequate to output a notification, detecting a loudness of a captured input signal, determining at least one of a loudness, a talking speed, and a mood of a speech signal within the captured input signal, detecting a duration of uninterrupted speech within the captured input signal, determining the loudness of a background signal comprised in the captured input signal, detecting a gap in the spectrum of a background signal comprised in the captured input signal, and determining information from at least one of a location sensor, calendar application, and phone settings.

Determining an adequateness parameter may comprise at least one of determining a high adequateness parameter if a speech pause has been detected in the situation information, determining a high adequateness parameter if a predefined speech sequence has been detected in the captured input signal which indicates from a human communication politeness aspect that it is currently adequate to output a notification, determining a low adequateness parameter if anything has been detected in the captured input signal that indicates from a human communication politeness aspect that it is currently not so adequate to output a notification, determining a low adequateness parameter if it has been detected that the input signal is loud, and determining a high adequateness parameter if it has been detected that the input signal is not loud.

Determining an adequateness parameter may also comprise determining a low adequateness parameter in situations where the surrounding is silent, but it is still inappropriate for the device to send out a notification.

Controlling the outputting of a notification may comprise at least one of adapting a notification signal, determining a mode for outputting the notification, determining whether or not the notification is to be output, and determining when the notification is to be output.

Adapting a notification signal may comprise adapting the volume, e.g. outputting the notification louder when the counterpart speaks loud.

Adapting a notification signal may also comprise adapting the talking speed of automatic speech output.

Adapting a notification signal may further comprise adapting automatic speech output so that it speaks faster when people speak fast.

Adapting a notification signal may also comprise adapting the voice of the automatic speech device to the mood of its users, i.e. speak enthusiastic if people speak enthusiastic.

Controlling the outputting of a notification may comprise selecting a mode for outputting the notification from at least one of speech output, a sound signal, a vibration alert and a visual signal.

Selecting a notification method can use one or more of the modalities as described in the following: In some situations, using a fixed audio file as notification mode is the right choice, e.g. when being alone in a silent environment. In this case, any sound recognizable as a notification signal will be appropriate. There are more situations where a fixed audio signal might be the right choice. E.g. in a cinema situation, a fixed audio signal that emulates a vibrating phone audio-wise might be more acceptable than a ringing bell or a voice.

Another possibility is to use a sound that normally naturally appears in the current situation of the user, but cannot be heard currently. In a cinema setting, e.g. the crinkling sound of a bag of chips is acceptable if not repeated too often.

Also, the opposite is possible: Using a natural sound that is atypical for the current situation. In an opera setting, a cricket sound might be acceptable or at least unsuspicious.

Any play-out of a fixed audio signal may be adapted for the suited point in time and the volume, as described above.

In some situations, the most adequate notification mode is using a spoken utterance. This can be the case e.g. in a cinema or opera situation when it is dark. In such a situation, a whispered statement is usual. By using a voice with attributes (like age and gender) not recorded around, the user can be informed without anyone else recognizing it as a disturbing electronic device.

In alternative embodiments, the automatic speech device is not repeating the same phrase over and over again, but varies the utterance.

Furthermore, when using voice, the user can be addressed with his/her (first) name or nickname in order to let the user know to whom the notification is sent.

Apart from varying the phonation mode (whispering vs. normal voice) in order to adapt the situation and from varying the gender and age of the speaker in order to avoid confusing the notification with a statement from another person, also the language and the frequency characteristics of a voice can be used to avoid this confusion.

Any play-out of a voice signal may be adapted for the suited point in time and the volume, as described above.

In other situations, using vibration as notification mode is an alternative choice. A vibration actuator has typically both a mechanical and an audible aspect: a vibrating device often can both be sensed on the skin and heard. Therefore, also the audible aspect of a vibrator can be managed in an adequate way.

In yet other situations, using visual signals such as flashing lights as notification mode is a good choice. A mobile device often features a display and other means to emit light. If the device can state that the device might be visible (e.g. by using a light sensor in the device that records a changing signal), also these means can be used in order to notify the user. Again, if a light measuring sensor exists, the "volume" (i.e. the intensity) of the light may be chosen in a way which is adequate.

The actual outputting of a notification can be shaped according to the operating environment and situation. For example, when the device knows the user is supposedly sitting in a meeting (by checking e.g. calendar entries), it could adjust to this situation by starting with a vibrating alert, or start muttering "pssst", or use similar, more human-like signaling. If the device is operated outside, it could use more aggressive methods, or if it is used in a concert hall it could only vibrate.

The context of the user may be sensed permanently. As soon as a notification needs to be done, a notification mode may be selected that fits the sensed context. The user then reacts either using the existing user interface means, e.g. by accepting or rejecting a call using buttons. Alternatively, the user may use a negative audio utterance (e.g. "shhht!" or "shut up") to stop the notification. In case of the negative audio utterance, the device assumes that the audio notification was not adequate. Therefore the device aims at using another alternative the next time.

Controlling the outputting of a notification may comprise at least one of adapting the notification to a detected loudness of the captured input signal, adapting the notification to at least one of a loudness, a talking speed, and a mood of a speech signal within the captured input signal, and adapting the notification to a detected loudness of a music signal comprised in the captured input signal.

The method for providing situation- adequate notifications may further comprise determining an urgency parameter related to a notification signal, and adapting the notification based on the urgency parameter.

When a speech device has to issue a notification it may try to get the floor, depending on the urgency of the notification. Less urgent notifications that are just for information such as "stay on the road for two miles" may only be issued when nobody else is talking.

In case of a more urgent command such as "turn right ahead" the device might wait for a short time to get the floor when people make a break, i.e. until a speech pause is detected in the captured input signal. If people continue to talk, the device might slowly get the attention by issuing a phrase such as "Please" and then wait for people interrupting their flow of speech. If this is not successful the attempt might be repeated with an increased volume.

Determining an urgency parameter may comprise increasing the urgency parameter with an increasing duration of uninterrupted speech within the captured input signal. Controlling the outputting of a notification may comprise comparing the adequateness parameter to the urgency parameter to determine the mode of outputting the notification.

Determining an urgency parameter may be based on at least on of content of a calendar entry, and a caller ID.

The method may further comprise decreasing a loudness of a music signal before outputting the notification signal. In this way the principle of "getting the floor" can also be extended to other situations, e.g. to the case that the person is listening to music at home. Before giving the desired speech output, the automatic speech device could first try to get the attention of the person (as described above) and then senses whether the music is turned down such that the person gives attention to the speech output that will follow. Furthermore, the device could be itself muting the music to get the attention of the person.

The concept of situation/ contextual awareness can also be used prophylactic. If a mobile device detects a situation where typically a loud ringing is not acceptable (like a meeting, a cinema, opera and the like), it can automatically create a vibration alarm after a few minutes. The alarm notifies the user (on a screen) that the device can either do nothing, or enter a so called intelligent nonobstructive notification mode, or switch off sound altogether (and rely on vibration alarm) or even switch off the device as a whole.

The user may also be able to prioritize, or exclude, certain notification methods in the settings. This may be helpful for users that don't want to use e.g. the Flashing Lights method, or prefer to use only the voice method. The user may also specify in what situations e.g. the voice method is adequate (i.e. in car, but not in meetings), and override or refine the learned presets.

Further, in a very noisy environment the device may detect gaps in the spectrum of the surrounding sounds and adjust the notification, so it can be heard even at a quite low level, because it can be easily distinguished from the background noise.

In addition to the described methods to "get the floor" and to shape the speech output in a situation- aware, non-obstructive way, the methods may use filler words or other means to indicate that the device is busy and to bridge processing time in a more adequate, natural way. The device also can focus on the important aspects only, if the situation suggests difficult communication conditions. And if the conditions are more relaxed, the device may supply more information compared to situations where time is scarce.

An electronic device for providing situation- adequate notifications comprises a situation recognizer module configured to determine situation information from characteristics of a captured input signal, an adequateness decider module configured to determine an adequateness parameter from the situation information, and a notification output module configured to control the outputting of a notification based on the adequateness parameter.

The electronic device may further comprise a notification source module configured to determine an urgency parameter related to the issuance of a notification signal, the notification output unit being further configured to control the outputting of a notification based on the urgency parameter.

The electronic device may further comprise a text-to-speech module, the text-to-speech module configured to provide, as a notification signal, a speech signal.

The electronic device may further comprise a sound database configured to provide, as a notification signal, a fixed audio signal.

The device may allow for more natural speech user interfaces and avoidance of awkward, machinelike communication. The speech output is used in a situation-adequate way and may take into account personal preferences, the environment and the situation when generating speech output. This may leads to less errors, misunderstandings, and may generally improve the acceptance of a speech user interfaces for the general consumer. For example, the device may be able to interrupt human speakers in a socially acceptable way in order to notify them, may learn how to interrupt human speakers in a socially acceptable way, may react to user utterances, e.g. in order to stop notifying a user, and may check whether the current notification settings match the current situation of the user and inform him (via a notification) about a potential mismatch between both.

Embodiments of devices and methods for providing situation-adequate notifications are now described in more detail with reference to the accompanying drawings.

Electronic device for providing situation-adequate notifications

Fig. 1 schematically shows an embodiment of an electronic device 1 for providing situation- adequate notifications. The electronic device 1 comprises a situation recognizer module 3 configured to determine situation information 29 from characteristics of a captured input signal 27, an adequateness decider module 5 configured to determine an adequateness parameter 31 from the situation information 29, and a notification output module 7 configured to control the outputting of a notification 33 based on the adequateness parameter 31 received from the adequateness decider module 5, and/ or based on the situation information 29 received from the situation recognizer module 3.

The situation recognizer module 3 is configured to receive an input signal 27 from different sensors, here a microphone 9, a GPS receiver 11, and a personal information retriever 13. The notification output module 7 is configured to issue a notification 33 via a loud speaker 15, a LED 17, or a vibration actuator 19. The notification output module 7 can thus choose between different output modes for outputting a notification.

The electronic device 1 further comprises a notification source module 21 configured to determine an urgency parameter 35 related to the issuance of a notification signal 33, the notification output unit 7 being further configured to control the outputting of a notification 33 based on the urgency parameter 35.

The electronic device 1 further comprises a text-to-speech module 23 configured to provide to the notification output unit 7, as a notification signal, a speech signal 37.

The electronic device 1 further comprises a sound database 25 configured to provide to the notification output unit 7, as a notification signal, a fixed audio signal.

All units and entities described in this specification and claimed in the appended claims can, if not stated otherwise, be implemented as integrated circuit logic, for example on a chip, and functionality provided by such units and entities can, if not stated otherwise, be implemented by software.

Method for providing situation-adequate notifications

Fig. 2 schematically shows an embodiment of a method for providing situation-adequate notifications. At S201, situation information is determined from characteristics of a captured input signal. At S203, an adequateness parameter is determined from the situation information. At S205, the outputting of a notification is controlled based on the adequateness parameter.

Getting the floor

Fig. 3 schematically shows an embodiment of a method for an automatic speech device trying to get the floor in order to output a speech notification.

At S301, it is determined from the situation information 29 whether or not a speech pause has been detected in the input signal 27. If yes, at S303, the speech notification is output.

If not, then, at S305, it is waited for some predefined time, e.g. 10 seconds, and, at S307 it is again determined from the situation information 29 whether or not a speech pause has been detected in the input signal 27. If yes, the attempt to get the floor was successful, and, at S303, the speech notification is output.

If not, i.e. there is no break for some time because people do not stop talking, at S309, a filler word such as "Please" or the first name of the person that the notification is addressed to is issued in order to interrupt politely. At S311 it is again determined from the situation information 29 whether or not a speech pause has been detected in the input signal 27. If yes, the attempt to get the floor was successful, and, at S303, the speech notification is output.

If not, i.e. there is again no break for some time because people do not stop talking, at S313, outputting the speech notification is started, e.g. by issuing the first two or three words of a complete phrase which is to be output. At S315 it is again determined from the situation information 29 whether or not a speech pause has been detected in the input signal 27. If yes, the attempt to get the floor was successful, and, at S303, the complete phrase output as speech notification.

If the attempt to get the floor was not successful, i.e. people do not stop talking, at S317, the speech notification is output with increased volume.

In order to implement this example method in an electronic device such as that of the embodiment of Fig. 1, the adequateness decider module 5 may be configured to determine a high adequateness parameter 31 if a speech pause has been detected in the situation information. The adequateness decider module 5 may further be configured to determine a low adequateness parameter 31 if no speech pause has been detected in the situation information. As a consequence of the adequateness parameter being low, the notification output module 7 may determine that it is not appropriate to issue a complete phrase as notification. Instead, the notification output module 7 may issue a filler word as described at S309 in Fig. 3, or the notification output module 7 might start with outputting the first two or three words of the complete phrase as described at S313 in Fig. 3. Should the notification output module 7 realize that the adequateness parameter 31 stays at a low level for a longer time, then it may choose to nevertheless output the complete sentence, eventually with increased volume as described at S317 in Fig. 3.

Ur ency of a notification

When a speech device has to issue a notification it may try to get the floor, depending on the urgency of the notification.

At S401, the urgency of a notification is determined. The urgency may for example be retrieved in the form of an urgency parameter 35 from notification source module 35 of the electronic device of Fig. 1.

At S403, it is determined whether a notification is urgent or not.

If it is determined at S403 that a notification is not urgent, i.e. in the case of a less urgent notification that is just for information such as "stay on the road for two miles", it is waited, at S405, for a speech pause. If a speech pause is detected, then, at S407, the speech notification is output. Thus, the less urgent notification is only issued when nobody else is talking.

If it is determined at S403 that a notification is urgent, e.g. in the case of more urgent notifications such as "turn right ahead" it is, at S409, waited for a short time to get the floor when people make a break, i.e. until a speech pause is detected in the captured input signal. If it is detected, at S411, that people continue to talk, it is slowly tried to get the attention by issuing, at S413, a filler word such as "Please" and then wait for people interrupting their flow of speech. If this is not successful the attempt might be repeated, at S417, with an increased volume. If this again is not successful, at S421, the complete speech notification is output with an increased volume.

In an alternative embodiment, determining an urgency parameter may comprise increasing the urgency parameter with an increasing duration of uninterrupted speech within the captured input signal.

In a further embodiment, controlling the outputting of a notification may comprise comparing the adequateness parameter to the urgency parameter to determine the mode of outputting the notification.

In a still further embodiment, determining an urgency parameter may be based on at least on of content of a calendar entry, and a caller ID.

***

It is thus disclosed in this specification:

(1) . An electronic device comprising

a situation recognizer module configured to determine situation information from characteristics of a captured input signal,

an adequateness decider module configured to determine an adequateness parameter from the situation information, and

a notification output module configured to control the outputting of a notification based on the adequateness parameter.

(2) . The electronic device of (1), further comprising a notification source module configured to determine an urgency parameter related to the issuance of a notification signal, the notification output unit being further configured to control the outputting of a notification based on the urgency parameter. (3). The electronic device of (1) or (2), further comprising a text-to-speech module, the text-to- speech module configured to provide, as a notification signal, a speech signal.

(4) . The electronic device of (1), (2), or (3), further comprising a sound database configured to provide, as a notification signal, a fixed audio signal.

(5) . A method comprising:

determining situation information from characteristics of a captured input signal, determining an adequateness parameter from the situation information, and

controlling the outputting of a notification based on the adequateness parameter.

(6) . The method of (5) in which controlling the outputting of a notification is based on the adequateness parameter and the situation information.

(7) . The method of (5) or (6), in which determining situation information comprises at least one of detecting a speech pause in the captured input signal,

detecting a predefined speech sequence in the captured input signal,

detecting anything that indicates from a human communication politeness aspect that it is currently adequate to output a notification,

detecting anything that indicates from a human communication politeness aspect that it is currently not so adequate to output a notification,

detecting a loudness of a captured input signal,

determining at least one of a loudness, a talking speed, and a mood of a speech signal within the captured input signal,

detecting a duration of uninterrupted speech within the captured input signal,

determining the loudness of a background signal comprised in the captured input signal, and detecting a gap in the spectrum of a background signal comprised in the captured input signal, and

determining information from at least one of a location sensor, calendar application, and phone settings.

(8) . The method of (5) to (7) in which determining an adequateness parameter comprises at least one of

determining a high adequateness parameter if a speech pause has been detected in the situation information, determining a high adequateness parameter if a predefined speech sequence has been detected in the captured input signal which indicates from a human communication politeness aspect that it is currently adequate to output a notification,

determining a low adequateness parameter if anything has been detected in the captured input signal that indicates from a human communication politeness aspect that it is currently not so adequate to output a notification,

determining a low adequateness parameter if it has been detected that the input signal is loud, and

determining a high adequateness parameter if it has been detected that the input signal is not loud.

(9) . The method of (5) to (8) in which the controlling the outputting of a notification comprises at least one of

adapting a notification signal,

determining a mode for outputting the notification,

determining whether or not the notification is to be output, and

determining when the notification is to be output.

(10) . The method of (5) to (9), in which controlling the outputting of a notification comprises determining a mode for outputting the notification from at least one of speech output, a sound signal, a vibration alert and a visual signal.

(11) . The method of (5) to (10) in which controlling the outputting of a notification comprises at least one of

adapting the notification to a detected loudness of the captured input signal,

adapting the notification to at least one of a loudness, a talking speed, and a mood of a speech signal within the captured input signal, and

adapting the notification to a detected loudness of a music signal comprised in the captured input signal.

(12) . The method of (5) to (11) further comprising outputting a notification signal which has the intention to interrupt the user.

(13) . The method of (5) to (12), further comprising determining an urgency parameter related to a notification signal, and adapting the notification based on the urgency parameter. (14). The method of (13), in which the determining an urgency parameter is based on at least on of content of a calendar entry, and a caller ID. (15). The method of (13) or (14), in which controlling the outputting of a notification comprises comparing the adequateness parameter to the urgency parameter to determine the mode of outputting the notification.

(16). The method of (13), (14), or (15), in which determining an urgency parameter comprises increasing the urgency parameter with an increasing duration of uninterrupted speech within the captured input signal.

(17). The method of (5) to (16), further comprising decreasing a loudness of a music signal before outputting the notification signal.

Claims

1. An electronic device comprising

2. The electronic device of claim 1, further comprising a notification source module configured to determine an urgency parameter related to the issuance of a notification signal, the notification output unit being further configured to control the outputting of a notification based on the urgency parameter.

3. The electronic device of claim 1, further comprising a text-to-speech module, the text-to- speech module configured to provide, as a notification signal, a speech signal.

4. The electronic device of claim 1, further comprising a sound database configured to provide, as a notification signal, a fixed audio signal.

5. A method comprising:

6. The method of claim 5 in which controlling the outputting of a notification is based on the adequateness parameter and the situation information.

7. The method of claim 5, in which determining situation information comprises at least one of

detecting a speech pause in the captured input signal,

detecting a predefined speech sequence in the captured input signal,

detecting anything that indicates from a human communication politeness aspect that it is currently adequate to output a notification, detecting anything that indicates from a human communication politeness aspect that it is currently not so adequate to output a notification,

detecting a loudness of a captured input signal,

detecting a duration of uninterrupted speech within the captured input signal,

8. The method of claim 5 in which determining an adequateness parameter comprises at least one of

determining a high adequateness parameter if a speech pause has been detected in the situation information,

determining a high adequateness parameter if a predefined speech sequence has been detected in the captured input signal which indicates from a human communication politeness aspect that it is currently adequate to output a notification,

9. The method of claim 5 in which the controlling the outputting of a notification comprises at least one of

adapting a notification signal,

determining a mode for outputting the notification,

determining whether or not the notification is to be output, and

determining when the notification is to be output.

10. The method of claim 5, in which controlling the outputting of a notification comprises determining a mode for outputting the notification from at least one of speech output, a sound signal, a vibration alert and a visual signal.

11. The method of claim 5 in which controlling the outputting of a notification comprises at least one of

adapting the notification to a detected loudness of the captured input signal,

12. The method of claim 5 further comprising outputting a notification signal which has the intention to interrupt the user.

13. The method of claim 5, further comprising determining an urgency parameter related to a notification signal, and adapting the notification based on the urgency parameter.

14. The method of claim 13, in which the determining an urgency parameter is based on at least on of content of a calendar entry, and a caller ID.

15. The method of claim 13, in which controlling the outputting of a notification comprises comparing the adequateness parameter to the urgency parameter to determine the mode of outputting the notification.

16. The method of claim 13, in which determining an urgency parameter comprises increasing the urgency parameter with an increasing duration of uninterrupted speech within the captured input signal.

17. The method of claim 5, further comprising decreasing a loudness of a music signal before outputting the notification signal.