JP2002169585A - Device and method for voice browser and storage medium with program stored therein - Google Patents

Abstract

PROBLEM TO BE SOLVED: To enable voice input even when a grammar dictionary for voice recognition is not prepared on the contents side. SOLUTION: The voice browser which is capable of voice input is equipped with a voice input part 107 which inputs a voice, a voice recognition part 108 which recognizes the inputted voice, a contents dictionary holding part 103 which holds a contents dictionary for voice recognition described in contents, a user dictionary holding part 105 which holds user dictionaries for voice recognition prepared by users, and a dictionary switching part 104 which switches a dictionary used by the voice recognition means to a user dictionary when a contents dictionary corresponding to an input field as an object of data input among input fields provided to the contents is absent.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

The present invention relates to a speech recognition method and apparatus suitable for a voice browser.

[0002]

2. Description of the Related Art With the spread of the Internet, Web information access has been generally performed. Furthermore, with the development of speech recognition / synthesis technology, Web
A so-called voice browser for accessing a page by voice has also been used (for example, Japanese Patent Application Laid-Open No. H11-249867).

[0003] Although the above describes accessing a Web page using only audio, a form in which a GUI and audio are used together when accessing a Web page is also considered (herein referred to as a composite browser). For example, JP-A-10-12
In 4293, a link in the Web page can be selected by voice input, and the contents of the Web page can be read out by synthesized voice. In this conventional example, by analyzing a word to which an anchor is attached and adding a reading, speech recognition of the word is possible. If the reading cannot be given, the user registers the reading to enable voice input.

In Japanese Patent Laid-Open No. Hei 10-124293, only a link can be selected by voice input, and input to an input field or the like is performed using a GUI. Here, if the content provider writes a grammar / dictionary for speech recognition in the content as is done in a general voice browser, the input to the input field and the like can be input by voice even in the complex type browser. It is easy to imagine what can be done. For example, in a virtual content description as shown in FIG.
The grammar / dictionary is indicated by the grammar attribute of the element. Here, the grammar / dictionary represented by “chimei” has, for example, as shown in FIG. Thus, the voice input to the input field 603 can be recognized.

[0005]

However, in the conventional complex type browser, if the grammar / dictionary is not described in the content, the voice cannot be recognized, and therefore, the input to the input field is performed by the voice. I couldn't do it. For this reason, only contents created by the content provider assuming voice input could be accessed by voice. For example, FIG.
Since no grammar / dictionary is described in the content shown in (1), no voice could be input to the input field 703.

The present invention provides a user dictionary for speech recognition, and performs speech recognition using the user dictionary even when a grammar / dictionary for speech recognition is not prepared on the content side. Is what makes it possible.

[0007]

In order to solve such a problem, for example, a voice browser device of the present invention has the following configuration. That is, a voice browser device capable of inputting by voice, comprising voice input means for inputting voice, voice recognition means for recognizing the input voice, and a content dictionary for voice recognition described in the content. A content dictionary holding means for holding, a user dictionary holding means for holding a user dictionary for voice recognition prepared for each user, and an input field which is a data input target among input fields provided in the content. A dictionary switching unit that switches a dictionary used by the voice recognition unit to a user dictionary when a corresponding content dictionary does not exist.

[0008]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS [Embodiment 1] Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.

FIG. 1 is a block diagram showing a basic configuration of an apparatus according to an embodiment of the present invention.

Reference numeral 101 denotes a content holding unit for holding content including content to be displayed by a browser.
Reference numeral 102 denotes a content analysis unit that analyzes the content held in the content holding unit 101. Reference numeral 103 denotes a content dictionary holding unit that holds a dictionary for speech recognition when the dictionary is described in the content. For example, data 4 as shown in FIG.
01 is retained. Reference numeral 104 denotes a dictionary switching unit that switches a dictionary for speech recognition used by the speech recognition unit 108. 10
Reference numeral 5 denotes a user dictionary holding unit that holds a user dictionary. The user dictionary holding unit 105 holds, for example, data 501 as shown in FIG. It is assumed that such data is registered in the user dictionary holding unit 105 in advance. 10
Reference numeral 6 denotes an input analysis unit that analyzes a voice input from the voice input unit 107 and a GUI input from the input unit 109. Reference numeral 107 denotes a voice input unit for performing voice input. Reference numeral 108 denotes a voice recognition unit that performs voice recognition. An input unit 109 performs a GUI input. A display unit 110 displays the content of the content. Reference numeral 111 denotes a content dictionary presence / absence determination unit that determines the presence / absence of a content dictionary.

FIG. 2 is a diagram showing a specific configuration of the apparatus of the present embodiment.

Reference numeral 201 denotes a CPU, which operates according to a program for realizing a procedure described later. Reference numeral 202 denotes a memory, which provides the content holding unit 101, the content dictionary holding unit 103, the user dictionary holding unit 105, and a storage area necessary for the operation of the program. Reference numeral 203 denotes a control memory, which stores a program that implements a procedure described later. Reference numeral 204 denotes a pointing device, and the input unit 1
09 is realized. Reference numeral 205 denotes a display, which implements the display unit 110. Reference numeral 206 denotes a microphone, which implements the voice input unit 107. A bus 207 connects the components.

Next, the operation of the apparatus of this embodiment will be described with reference to the flowchart shown in FIG. In the present embodiment, a case where one input field exists in one content is handled.

First, in step S301, the network
The content is acquired from a device (not shown) or the like, and is stored in the content storage unit 101.

In step S302, the contents of the contents held in the contents holding unit 101 are analyzed by the contents analyzing unit 102. If the grammar / dictionary for speech recognition is described in the content, the grammar / dictionary is extracted based on the description. It is assumed that the grammar / dictionary is described in the content by tags and attributes of the content description language. For example, in the virtual content description of FIG. 6, since the file name is described by the grammar attribute of the input element, the file is acquired.

In step S303, step S302
The grammar / dictionary for speech recognition extracted in step (1) is stored in the content dictionary storage unit 103.

In step S304, the input analysis unit 106 checks whether an input has been made. This step is repeated until there is an actual input.

In step S305, step S304
Check if the input in is a voice input into the input field. Step S for voice input into the input field
Proceed to 306. If not, the process ends.

In step S306, if the automatic mode is selected in the mode for selecting a dictionary to be used in the voice recognition unit, the flow advances to step S307.

In step S307, it is determined whether the content dictionary is held or not by the content dictionary presence / absence determining unit 1.
Check at 11. If the content dictionary is stored, the process proceeds to step S308. If the content dictionary is not stored, the process proceeds to step S309.

In step S308, the dictionary switching unit 104
In, the dictionary used by the voice recognition unit is switched to the user dictionary stored in the user dictionary storage unit 105.

In step S309, dictionary switching unit 104
, The dictionary used by the voice recognition unit 108 is switched to the content dictionary stored in the content dictionary storage unit 101.

In step S310, step S304
Performs voice recognition processing of the voice input input by. If there is a content dictionary, the content dictionary stored in the content dictionary storage unit 103 is used. If there is no content dictionary, the user dictionary stored in the user dictionary storage unit 105 is used.

In step S311, step S308
The result of the voice recognition is displayed on the display unit 110. Then, the process ends.

Next, the processing procedure of this embodiment will be further described using a specific example.

If a part of the content is as shown in FIG. 6, a grammar / dictionary is specified in the content.
(grammar attribute of the input element), so that it is acquired and stored in the content dictionary storage unit 103. Then, when a voice is input in the input field 603,
Voice recognition is performed using this dictionary (for example, 401). On the other hand, if a part of the content is as shown in FIG.
The grammar / dictionary is not specified in the content. Therefore, in the conventional example, it is not possible to input voice into the input field 703. In this embodiment, in this case, in such a case, the user dictionary (for example, 5
01), the voice can be input to the input field 703 by voice.

[Embodiment 2] In the above embodiment, when the grammar / dictionary is not described in the content, the user dictionary is automatically switched to the user dictionary in step S307 in the automatic mode. The user dictionary may be used when explicitly specified by the user. In addition to the configuration of FIG. 1, means for designating the use of the user dictionary may be provided, and the dictionary may be switched when the user designates the dictionary. That is, when the manual (user) mode is selected in step S306, the process proceeds to step S308, and the user dictionary is always used.

[Embodiment 3] In addition, according to the same configuration as the above-mentioned other embodiments, even when a dictionary is described in the content, a user dictionary can be used according to a user's designation. It is possible. For example, when a part of the content is as shown in FIG. 6, a grammar / dictionary is specified in the content (for example, the dictionary in FIG. 4). Here, when the user explicitly specifies the use of the user dictionary, when a voice is input to the input field 603, the user does not use the content dictionary but the user dictionary (for example, FIG. 5).
Speech recognition is performed using the That is, when the manual (user) mode is selected in step S306, the process proceeds to step S308, and the user dictionary is always used. When the manual (content) mode is selected in step S306, the process proceeds to step S309, and it is possible to set so that the content dictionary is always used.

[Embodiment 4] In the above embodiment, the case where there is one user dictionary has been described, but there may be a plurality of user dictionaries. As to which one of the plurality of dictionaries to use, there is a method of explicitly specifying the dictionary or a method of automatically determining the dictionary.

The former can be carried out in the same manner as the other embodiments.

The latter can be implemented, for example, as follows. A classification category (representing a field related to which field) is assigned to each user dictionary in advance. It is determined which classification category the content belongs to (this may be done by a known document classification method), and a user dictionary of the corresponding classification category is adopted. A user dictionary to be used when a user dictionary of the corresponding category does not exist may be determined.

[Fifth Embodiment] In the above embodiment, the case where one input field is present is taken as an example, but a plurality of input fields may exist. This case can be dealt with by performing the same processing as in the above embodiment for each input field.

[Embodiment 6] In the above-described embodiment, the case where the contents of the content dictionary and the user dictionary are simple dictionaries as shown in FIGS. 4 and 5 is taken as an example. The recognition grammar can be similarly implemented.

[Embodiment 7] In the above embodiment, the audio output of the content is not performed. However, the same can be applied to the case where the audio output of the content is performed.

[Embodiment 8] In the above embodiment, the case where each unit is configured on the same computer has been described. However, the present invention is not limited to this, and may be realized on a plurality of computers.

[Embodiment 9] The present invention may be applied to a system composed of a plurality of devices or an apparatus composed of one device. A recording medium storing software program codes for realizing the functions of the above-described embodiments is supplied to a system or an apparatus, and a computer (or CPU or MP) of the system or the apparatus is provided.
Needless to say, this can also be achieved by U) reading and executing the program code stored in the recording medium.

In this case, the program code itself read from the recording medium implements the functions of the above-described embodiment, and the program code itself implements the functions of the above-described embodiment. The recording medium on which the code is recorded constitutes the present invention.

[Embodiment 10] As a recording medium for supplying a program code, for example, a floppy (registered trademark) disk, hard disk, optical disk, CD-
A ROM, a CD-R, a DVD-ROM, a magnetic tape, a nonvolatile memory card, a ROM, and the like can be used. [Embodiment 11] When the computer executes the readout program code, not only the functions of the above-described embodiment are realized, but also an OS running on the computer based on the instruction of the program code. Performs part or all of the actual processing, and the processing realizes the functions of the above-described embodiments.

[Twelfth Embodiment] Further, after the program code read from the recording medium is written into a memory provided in a function expansion board inserted into the computer or a function expansion unit connected to the computer, the program code is read. It is needless to say that a CPU or the like provided in the function expansion board or the function expansion unit performs part or all of the actual processing based on the instruction, and the function of the above-described embodiment is realized by the processing. .

[0040]

As described above, according to the present invention,
Even in a content in which a grammar / dictionary for voice recognition is not described in the content, the content can be input by voice.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a basic configuration of an embodiment of a voice browser device according to the present invention.

FIG. 2 is a diagram showing a specific configuration of an embodiment of the present invention.

FIG. 3 is a flowchart showing an outline of processing in the embodiment of the voice browser device according to the present invention.

FIG. 4 is a diagram showing an example of contents of a content dictionary according to the embodiment of the present invention.

FIG. 5 is a diagram illustrating an example of contents of a user dictionary according to the embodiment of the present invention.

FIG. 6 is a diagram for specifically describing an embodiment of the present invention.

FIG. 7 is a diagram for specifically describing an embodiment of the present invention.

[Explanation of symbols]

 101: Content holding unit 102: Content analyzing unit 103: Content dictionary holding unit 104: Dictionary switching unit 105: User dictionary holding unit 106: Input analyzing unit 107: Voice input Unit 108: Voice recognition unit 109: Input unit 110: Display unit 111: Content dictionary presence / absence determination unit 201: CPU 202: Memory 203: Control memory 204: Pointing Device 205 ・ ・ ・ Display 206 ・ ・ ・ Microphone 207 ・ ・ ・ Bus

──────────────────────────────────────────────────続 き Continued on the front page (51) Int.Cl. ⁷ Identification code FI Theme coat ゛ (Reference) G10L 15/00 G10L 3/00 551P 15/28 F term (Reference) 5B075 KK07 KK13 KK33 KK37 ND03 ND14 ND20 ND23 ND36 NK34 PP07 PP12 PP30 PQ02 PQ04 PQ42 UU40 5D015 KK01 LL10

Claims (7)

[Claims] 1. A voice browser device capable of inputting by voice, comprising: voice input means for inputting voice; voice recognition means for recognizing the input voice; and voice recognition means for voice recognition described in the content. Content dictionary holding means for holding a content dictionary, user dictionary holding means for holding a speech recognition user dictionary prepared for each user, and data input targets among input fields provided in the content A voice browser device comprising: a dictionary switching unit that switches a dictionary used by the voice recognition unit to a user dictionary when a content dictionary corresponding to the input field does not exist. 2. The voice browser device according to claim 1, further comprising a content presence / absence determining means for determining the presence / absence of a content dictionary, and automatically switching to a user dictionary when the content dictionary does not exist. 3. The voice browser device according to claim 1, further comprising a user dictionary designating means for explicitly instructing the user to use the user dictionary, and switching to the user dictionary according to the user's instruction. 4. A voice browser method capable of inputting by voice, comprising: a voice inputting step of inputting a voice; a voice recognizing step of recognizing the input voice; A content dictionary holding step of holding a content dictionary, a user dictionary holding step of holding a voice recognition user dictionary prepared for each user, and data input targets among input fields provided in the content. A dictionary switching step of switching a dictionary used in the voice recognition step to a user dictionary when there is no content dictionary corresponding to the input field. 5. The voice browser method according to claim 4, further comprising a content presence / absence determination step of determining presence / absence of a content dictionary, and automatically switching to a user dictionary when the content dictionary does not exist. 6. The voice browser method according to claim 4, further comprising a user dictionary specifying step in which the user explicitly instructs use of the user dictionary, and switching to the user dictionary according to the user's instruction. 7. A computer-readable storage medium storing the steps included in the voice browser method according to claim 4 as a program for causing a computer to execute the steps.