DE102010054242A1 - Method for providing operating device for operating telephone device mounted in vehicle, involves determining input sequence fragment indicating continuation and completion of voice input, and providing graphic object for fragment - Google Patents

Abstract

The method involves detecting voice input with input sequence. The graphic object for input sequence is displayed in vehicle. The input sequence fragment indicating continuation and completion of voice input is determined. The graphic object for input sequence fragment is provided. Independent claims are included for the following: (1) operating device; and (2) vehicle.

Description

The present invention relates to a method for providing an operating device in a vehicle, in which a voice input is detected, by the detected voice input, an input sequence is initiated or continued, and a graphical object associated with the input sequence is displayed in the vehicle. The present invention further relates to an associated operating device in a vehicle.

Operating devices that can be operated at least partially via voice inputs are increasingly being used in vehicles. As a result, the ease of use of the numerous vehicle devices is improved, such as the operation of a navigation system, a telecommunication device or an infotainment device. The operating concept can allow user input of various modalities. In particular, multi-modal voice input devices are known with a display device, wherein user inputs are stored in the form of an input sequence and rendered graphically by the display device. The user can then make further input via speech or via an interaction with the graphical representation. An input sequence is finally assigned a control command, by the triggering of which the associated vehicle function is executed. The recorded user inputs are first converted into a data-technical digital signal and assigned by an algorithm also a data-technically stored control command, which in turn is usually converted via an electrical control pulse into a physical signal. By means of various transducers, it is thus possible in particular to control mechanical, acoustic or optical output variables of the various vehicle devices.

The amount of information and the variety of functions in vehicles has risen sharply in recent years. In addition to a steadily increasing number of data stored technically operating menus for controlling various vehicle equipment, such. As navigation system or air conditioning, especially the variety of functions in modern infotainment facilities has risen sharply. The user must be able to unerringly find or trigger a file or function in the vehicle even without knowledge of the information architecture. This is particularly important if the user is the driver of the vehicle so that it is not distracted from the traffic by the operating actions.

The DE 103 60 656 A1 relates to a voice input control system for a vehicle. In this case, by means of the operating context, which is determined by the already made selection of a menu level by manual operating means, when opening the voice interface, a language dialogue associated with this menu level is opened. If several possible operating intentions are detected, keywords for the various operating actions are displayed for selection.

The DE 196 46 634 A1 describes a voice interface in which an input word sequence is assigned to a stored voice command and this is displayed graphically. This can be confirmed or rejected by the user. In case of ambiguity, a selection list can be displayed from which the user can then select the desired entry.

It is the object of the present invention to provide a method for providing an operating device of the aforementioned type and an associated operating device, by means of which a quick and intuitive operation is made possible.

This object is achieved by a method having the features of claim 1 and an operating device with the features of claim 9. Advantageous training and further developments emerge from the dependent claims.

The method according to the invention is characterized in that one or more alternative input sequence fragment (s) is automatically determined depending on the input speech sequence associated with the input sequence, wherein an input sequence fragment comprises a possible continuation or completion of the detected speech input. According to the invention, the graphical object associated with the input sequence further comprises these input sequence fragments. By exploiting incomplete voice commands for voice control, the user interaction may be shortened or accelerated if a user z. B. was interrupted during voice input or comes to a halt. The user is then given the opportunity to complete the voice input quickly and intuitively. The graphical display helps him discover the operable functions and learn valid commands based on a prediction derived from the previously acquired speech input and / or input sequence.

The method according to the invention furthermore makes it possible to reduce the source of error that the user makes incorrect or inappropriate or intended voice inputs. This is especially important in a vehicle because it is there ambient noise causes further sources of error which can make correct speech recognition more difficult. By reducing the number of independent sources of error, the speech dialogue becomes more efficient and prevents the driver of the vehicle from being distracted from the traffic.

For example, possible and / or probable continuations or completions are displayed in a hit list. The hit list may have one or more input sequence fragments as entries or may be empty. Conveniently, a longer list of typically more than three to five entries may be sorted by relevance since all entries would no longer be visually detectable by the user at a glance. If no matching entry is found, either an empty list can be output as a result or a list of the next best, possibly irrelevant, hits. The relevance can be calculated according to various known algorithms.

In particular, it may be provided that the input sequence is extended by a selected input sequence fragment, wherein an input sequence fragment is automatically selected when only one input sequence fragment has been determined, or is selected by a user input from a plurality of alternative input sequence fragments. A user input may in this case be another voice input. In particular, however, it may also be a manual input by the user, for example by selecting an input sequence fragment displayed on a display area, e.g. By tapping an input sequence fragment displayed on a touch screen. The non-selected input sequence fragments are then typically discarded, but may be cached for later use.

Speech inputs and / or input sequence fragments exist for. B. from a whole word. Firstly, these are easy to memorize on the part of the user and, secondly, can also be detected visually quickly if they are displayed by means of a graphic object. However, input sequence fragments can also be formed from sentence fragments, e.g. B. short half sentences, standing words or multi-word proper names.

The method is capable of being carried out iteratively, that is to say that multiple voice inputs are detected successively and associated input sequence fragments are selected. The display of the graphical object is expediently updated after each user input, in particular after every voice input. In this way, even longer input sequences can lead to a fast and targeted user input.

In particular, the method can be continued iteratively until the input sequence represents a complete command and thus no input sequence fragments are found.

During the iterative execution, it may happen that an input sequence fragment has already been automatically selected on the system side, but a part of this input sequence fragment is repeated in the subsequent speech input by the user. For example, a less skilled user is unaware of automatic selection of unique input sequence fragments. In this case, the method may provide for deleting the unnecessarily spoken words from the input sequence, or not even recording them at all.

In order to determine the input sequence fragments, for example, a detected speech input and / or the input sequence are analyzed grammatically and / or compared with stored speech commands. It is z. For example, it evaluates which part of speech is likely to be in the previous input sequence or which attributes match a stored voice command. This can increase the speed of finding relevant, probable or possible input sequence fragments. Thus, not only static commands but also possible input sequences are displayed in the graphical object.

If an input sequence is uniquely assigned to a voice command, it can be executed directly and automatically. Alternatively, however, a graphical output may appear again to obtain confirmation from the user as to whether this command should actually be executed. In any case, it can be provided that a successful selection of a voice command is displayed graphically.

In one embodiment of the method according to the invention, the voice commands control functional devices in a vehicle. In particular, so that infotainment devices, navigation systems or about the air conditioning in the vehicle can be conveniently operated. The mixed approach of getting a desired input sequence as quickly as possible by means of voice inputs and selected input sequence fragments helps the driver of a vehicle not be distracted from the traffic when operating functional devices.

The operating device according to the invention comprises means for detecting speech inputs, and a memory unit for storing an input sequence that is initiatable or continuable by detected speech inputs. It further comprises a display area for displaying a graphic object associated with the input sequence and a control device connected to the means for detecting speech inputs, the storage unit and the display area. It is characterized in that, by means of the control device, one or more alternative input sequence fragment (s) is automatically detectable depending on the detected speech input associated with the input sequence, wherein an input sequence fragment comprises a possible continuation or completion of the detected speech input, and the display surface such is controlled by the control unit that the graphic object associated with the input sequence further comprises these input sequence fragments. The device is particularly suitable for carrying out the method according to the invention. It therefore also has the advantages of the method according to the invention.

Furthermore, according to the invention, a vehicle is equipped with such an operating device.

The invention will now be explained in more detail by means of embodiments with reference to the figures.

1 shows schematically the structure of an embodiment of the operating device according to the invention,

the 2a - 2c schematically show user inputs and the determination of input sequence sequences in the course of an iteratively performed method according to an embodiment of the invention, and

the 3a - 3c show schematically graphic objects that in the course of the iterative method according to the in the 2a - 2c shown embodiment display the input sequences.

In 1 schematically the structure of an embodiment of the operating device according to the invention is shown. A display area 1 in a vehicle is via a control unit 4 with one or more manual controls 2A connected via which a user commands or inputs for the display area 1 can make. The display area 1 is a freely programmable display area, eg. B. a per se known TFT or OLED display. The manual control 2A may be, for example, a rotary trigger or a so-called touch screen. On the display surface 1 Various information can be graphically displayed. In particular, the information is a graphic representation of a speech dialogue in the form of a graphic object 10 , which will be described in detail below.

The control unit 4 is further with funds 3 for detecting a voice input and the data bus 6 connected in the vehicle through which they access one or more databases 7 . 7A . 7B Has. This can be a cross-database 7 be in the all accumulated in the vehicle data and data stored control commands of various vehicle equipment 8A . 8B are available. Every vehicle interior 8A . 8B can also have its own database 7A . 7B in which the to the respective vehicle device 8A . 8B specific data and data stored control commands are stored. There are two vehicle facilities by way of example 8A . 8B but it can also be just one or a plurality of such devices. It is, for example, a telephone system, a navigation system, an infotainment device or the air conditioning system in the vehicle.

The means 3 for detecting a voice input include a microphone for receiving acoustic signals, a software module for voice activity detection and an adaptive filter for noise reduction. The voice activity detection and the adaptive filter may alternatively be located in other hardware and, for example, via the control unit 4 with the means 3 to capture the speech input. The display area 1 is in the example shown wholly or partially with a touch-sensitive surface 2 B equipped so that over the display area 1 manual entries can be made. For example, the user could by touching the appropriate location on the display area 1 Select certain parts of the graphical display, as will be explained in more detail below in connection with the inventive method.

The control unit 4 includes a storage unit 5 for storing input sequences 14 and for storing a hit list, the input string fragments 15 contains one or more possible continuations or completions of a recorded speech input 20 include.

The inventive method will now be described with reference to an embodiment referred to as a word-based "speller" with reference to 2a - 2c such as 3a - 3c explained in more detail. Here are in the 2a - 2c schematic user input 20 . 21 and the determination of input sequence sequences 15 in the course of an iterative carried out procedure. The corresponding schematic display of the graphical object 10 is in the 3a - 3c shown. For carrying out the method, in particular those relating to the 1 described operating device can be used.

For example, a user would like to make the voice input: "Start Navigation to Berlin". He initiates with a voice input 20 a graphically supported speech dialog and starts with the command word "Start" as the first voice input 20 at. There may now be grammatically several meaningful commands for starting a function or user application, eg. As the start of the radio, an MP3 player or other media player, or a navigation program.

The voice input 20 is recognized as a whole word, grammatically analyzed and included in the various databases 7 . 7A . 7B stored voice commands compared. This is the voice input 20 as the beginning of an input sequence 14 in the storage unit 5 saved. To the input sequence 14 will automatically after input sequence fragments 15 searched for the voice input 20 fit and thus possible continuations or completions of the recorded speech input 20 represent. These continuations or completions belong to a free speech input and not necessarily to a particular static voice command.

The Indian 2a represented dialogue tree of possible subsequent words or continuations of the expected user-side command input can by means of a comparison with the databases 7 . 7A . 7B be determined. In the specific case, for example, the three previously mentioned functional devices in the vehicle as input sequence fragments 15 found and then as whole words in the graphic display 12 together with the graphic display 11 the stored input sequence 14 as a graphic object 10 displayed as shown in the 3a is shown. As a result, the user can not only learn the correct commands expected by the system, but also z. B. learn about it not known operating options.

In the following step ( 2 B ), the user can now either by means of another voice input 20 say the system-correct term "navigation" or this by a user input 21 about another modality, eg. B. by touching the corresponding display 12 the corresponding input sequence fragment 15 , enter manually. In particular, another voice input 20 the best option for the driver, the display 12 the input sequence fragments 15 but he supports it visually.

The visual support increases the likelihood of correct command input. Thus it can be more reliably assumed in the case of a user input that a word which was valid for the relevant voice command was actually spoken. Thus, speech recognition is particularly simplified in a background noise environment, as is often the case in a vehicle, since an independent source of error can be eliminated or at least reduced.

If the system recognizes that only one possible input sequence fragment is required for an input sequence 15 ' This is automatically selected in the exemplary embodiment and, in a further iteration step, the next input sequence fragments 15 certainly. In concrete example, it is recognized that the word "navigation" can only follow the word "after".

According to the iterative traversal, the displayed graphic object becomes 10 Updated as in the 3b is shown. It contains the updated ad 11 the input sequence 14 also the automatically selected input sequence fragment 15 ' , Furthermore, the ad 12 the newly determined input sequence fragments 15 in the graphic object 10 updated.

In 2c is shown as the user through another voice input 20 completes the voice command by entering the navigation destination "Berlin". The display of the graphical object 10 is updated again as shown in the 3c is shown.

Became the input sequence 14 As long as extended until a complete voice command or a complete voice input results, this or an associated voice command can be executed automatically without delay. To indicate this to the user, in addition to the graphical object 10 to the now complete input sequence 11 another affirmation object 13 which indicates to the user that his voice command has been correctly recorded. Such an affirmation object 13 can consist of a confirmation hook, but can also, for example, go into an animation, especially if the display area 1 on which the graphic object 10 is now required for the execution of the vehicle function. For example, an animated route plan for navigation to Berlin is displayed.

LIST OF REFERENCE NUMBERS

1

display area

2A

manual control

2 B

touch-sensitive surface

3

Means for detecting speech input

4

control device

5

storage unit

6

Data bus in the vehicle

7, 7A, 7B

databases

8A, 8B

Functional devices in the vehicle

10

graphic object

11

graphical display of the input sequence

12

graphical display of the input sequence fragment

13

Affirmationsobjekt

14

input sequence

15, 15 '

Input sequence fragment

20

voice input

21

manual user input

QUOTES INCLUDE IN THE DESCRIPTION

This list of the documents listed by the applicant has been generated automatically and is included solely for the better information of the reader. The list is not part of the German patent or utility model application. The DPMA assumes no liability for any errors or omissions.

Cited patent literature

• DE 10360656 A1 [0004]
• DE 19646634 A1 [0005]

Claims (10)

Method for providing an operating device in a vehicle, in which - a voice input ( 20 ), - by the recorded speech input ( 20 ) an input sequence ( 14 ) is initiated or continued, and - an input sequence ( 14 ) belonging graphic object ( 10 ) is displayed in the vehicle, characterized in that - depending on the input sequence ( 14 ) recorded speech input ( 20 ) automatically one or more alternative input sequence fragments (s) ( 15 ), where an input sequence fragment is a possible continuation or completion of the acquired speech input ( 20 ), and - the input sequence ( 14 ) associated graphic object ( 10 ), these input sequence fragments ( 15 ). Method according to claim 1, characterized in that the input sequence ( 14 ) around a selected input sequence fragment ( 15 ; 15 ' ), where an input sequence fragment ( 15 ' ) is automatically selected if only one input sequence fragment ( 15 ' ) or by a user input ( 20 ; 21 ) from several alternative input sequence fragments ( 15 ) is selected. Method according to claim 1 or 2, characterized in that the speech input ( 20 ) and / or an input sequence fragment ( 15 ) consists of a whole word. Method according to one of claims 1 to 3, characterized in that the method is carried out iteratively and the display of the graphical object ( 10 ) after each user input ( 20 ; 21 ) is updated. Method according to one of the preceding claims, characterized in that for determining the input sequence fragments ( 15 ) a recorded speech input ( 20 ) and / or the input sequence ( 14 ) is grammatically analyzed and / or compared with stored voice commands. Method according to one of the preceding claims, characterized in that a to the input sequence ( 14 ) uniquely assigned voice command is executed immediately. A method according to claim 6, characterized in that a successful selection of a voice command is displayed graphically ( 13 ). Method according to claim 6 or 7, characterized in that by the voice commands functional devices ( 8A . 8B ) are controlled in a vehicle. Operating device in a vehicle, comprising - means ( 3 ) for detecting voice inputs, - a memory unit ( 5 ) for storing an input sequence ( 14 ) detected by recorded voice input ( 20 ) can be initiated or continued, - a display area ( 1 ) to display an input sequence ( 14 ) belonging graphic object ( 10 ) and - a control device ( 4 ), with the means ( 3 ) for detecting voice input, the memory unit ( 5 ) and the display area ( 1 ), characterized in that - by means of the control device ( 4 ) depending on the input sequence ( 14 ) recorded speech input ( 20 ) automatically one or more alternative input sequence fragments (s) ( 15 ) is detectable, wherein an input sequence fragment a possible continuation or completion of the detected speech input ( 20 ), and - the display area ( 1 ) from the control unit ( 4 ) that the input sequence ( 14 ) associated graphic object ( 10 ), these input sequence fragments ( 15 ). Vehicle with an operating device according to claim 9.

Images