Data processing: speech signal processing – linguistics – language – Speech signal processing – Application
Reexamination Certificate
2000-06-30
2004-10-05
Knepper, David D. (Department: 2654)
Data processing: speech signal processing, linguistics, language
Speech signal processing
Application
C704S251000
Reexamination Certificate
active
06801896
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates to a voice-based manipulation technique capable of controlling and manipulating electronic devices or the like through input voices, and, more particularly, to a voice-based manipulation method and apparatus, which allow even a user who does not remember registered words to easily check the correlation between registered words and subjects to be manipulated, thereby improving the operability.
2. Description of the Related Art
Voice-based manipulation techniques which permit a user to manipulate electronic devices or the like through input voices have been proposed. Meanwhile, some improvements on speech recognition techniques have also been made. With such improved speech recognition techniques, there are active developments of electronic devices or the like which use voice-based manipulation techniques.
For example, there is known an on-board audio system for a vehicle, which can be manipulated with voices in the following manner. Using this audio system, a user registers voice data for each of the channel frequencies of broadcasting stations. When the user utters some words corresponding to one of the registered voice data, the audio system recognizes the uttered words through a speech recognition technique and automatically tunes itself to the designated channel frequency.
More specifically, the user tunes in to the channel frequency of a desired broadcasting station and utters words, for example, “first broadcasting station”, by manipulating a voice registration button provided on the on-board audio system, and then voice data of the words “first broadcasting station” can be stored (registered) in a memory in association with that channel frequency. In a similar fashion, the user tunes the audio system to the channel frequencies of other broadcasting stations and utters words, such as “second broadcasting station” and “third broadcasting station”. As a result, voice data of the words “second broadcasting station”, “third broadcasting station” and so forth can be stored in the memory in association with the tuned channel frequencies. When, after this voice registering operation, the user utters one stream of words, selected from the registered groups of words, such as “first broadcasting station”, “second broadcasting station” and “third broadcasting station”, the audio system recognizes the voiced words and automatically tunes itself to the designated channel frequency.
As mentioned above, this on-board audio system can permits voice-based manipulation to be performed based on voice data that has been registered beforehand in association with subjects to be manipulated (hereinafter referred to as “manipulation targets”). But, users are likely to forget registered words or forget the correlation between the registered words and manipulation targets. In this case, each user may have to, for example, repeat the above-described voice registering operation to change old voice data stored in the memory to new voice data.
It is desirable to ensure voice registration of any words, not specific words, thereby improving the operability for users. If such a highly general-purpose design is taken, the designed audio system, though effective in many ways, would suffer a lower operability, because users are apt to forget registered words.
While the tuning operation of an on-board audio system has been specifically discussed to show the problem of the conventional voice-based manipulation techniques, the same problem arises in the case where a user who is likely to forget registered words loads a recording/reproducing medium in an MD (Mini Disc) player, CD (Compact Disc) player or the like, which is installed in an on-board audio system and selects a musical piece, a title or the like, recorded on that medium, with voices.
The factor that users may forget registered words is the problem that should be overcome not only for on-board audio systems for vehicles but also the voice-based manipulation techniques.
SUMMARY OF THE INVENTION
Accordingly, it is an object of the present invention to provide a voice-based manipulation method and apparatus, which allow even a user who has forgotten registered words to easily check the correlation between registered words and manipulation targets, thereby ensuring an improved operability.
To achieve the above object, according to one aspect of this invention, there is provided a voice-based manipulation apparatus which comprises a storage section for storing voice information for specifying manipulation targets in association with the manipulation targets; a manipulation section for, when a voice is supplied, manipulating the manipulation targets which is associated with the voice information stored in the storage section which corresponds to the voice; and a search section for searching the voice information stored in the storage section in association with the manipulation target and presenting the resultant voice information.
According to another aspect of this invention, there is provided a voice-based manipulation method which comprises the steps of storing voice information for specifying manipulation targets in a storage section in association with the manipulation targets; manipulating, when a voice is supplied, the manipulation targets which corresponds to the voice; and searching the voice information stored in the storage section in association with the manipulation target and presenting resultant voice information.
With the above structures, a user can acquire voice information which is searched and presented by the search section (or the searching step). Even if the user forgets, or is uncertain abut, voice information stored (registered) in the storage section, the user can easily check the correlation between the voice information and the manipulation target which is associated with that voice information. Even when the user does not remember voice information, therefore, it is unnecessary to store voice information again in the storage section, resulting in an improved operability.
It is preferable in the above voice-based manipulation apparatus and method that, in response a search instruction externally supplied, the search section or the searching step should detect an active manipulation target, search for that voice information which is associated with the detected active manipulation target and present the searched voice information.
In this case, when voice information associated with the active manipulation target is not stored in the storage section, the search section or the searching steps may search other voice information stored in the storage section in association with the manipulation target and present the searched voice information.
In the above two preferable modes, it is further preferable that in response to the search instruction externally supplied, the search section or the searching step should search the voice information stored in the storage section in a predetermined order in association with the manipulation target and present the searched voice information.
In this case, the predetermined order may be an alphabetical order, a forward sort direction or a reverse sort direction.
In the voice-based manipulation apparatus according to the first aspect, the voice-based manipulation method according to the second aspect, or any one of the above-described preferable modes, the storage section can store the voice information again and may store a supplied voice as voice information associated with an active manipulation target at the time of storing the voice information again.
REFERENCES:
patent: 5257314 (1993-10-01), Kimura
patent: 5267323 (1993-11-01), Kimura
patent: 6507815 (2003-01-01), Yamamoto
ArentFox
Knepper David D.
Pioneer Corporation
LandOfFree
Voice-based search and selection of speech recognition data does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Voice-based search and selection of speech recognition data, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Voice-based search and selection of speech recognition data will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3296565