JP2001282285A5 - - Google Patents

Description

0003
FIG. 5 shows a configuration diagram of a conventional voice recognition device, which will be described below.
Before starting voice recognition, a plurality of words to be voice recognition are divided into a plurality of predefined genres and stored as a word set 50, and the dictionary generation unit 51 extracts the voice features of the plurality of word sets 50. Then, it is stored in the dictionary file 52, and the dictionary loading unit 53 loads the instructed dictionary file 52 of the dictionary loading instruction unit 54 into the dictionary memory 55 before voice input.

0005
[Problems to be Solved by the Invention]
However, even if these conventional techniques are used, the vocabulary to be recognized is not always intended by the user because it is created according to a predetermined genre, and the dictionary file is loaded every time the predetermined genre changes. It is necessary to repeat the above, and there are problems such as an increase in user operations in order to obtain a voice recognition result that matches the user's intention.

0006
An object of the present invention is to improve the speech recognition performance by narrowing down the recognition target vocabulary according to the user's preference with a small number of user operations.

0010
BEST MODE FOR CARRYING OUT THE INVENTION [Embodiments of the Invention]
In the invention according to claim 1 of the present invention, a recognition target vocabulary based on user preference information acquired in advance from a recognition target candidate vocabulary storing all words to be voice recognition is created and input in advance. It is characterized by recognizing a word from a voice and a recognition target vocabulary, and has an effect of improving the voice recognition performance by narrowing down the recognition target vocabulary according to the user's preference. In addition, it is a method that is simpler and faster than performing speech recognition for all vocabularies that can be recognition target candidates.
The invention according to claim 2 is based on a recognition score obtained by voice recognition for recognizing the word from a input voice and a recognition target vocabulary which is a word to be voice-recognized, and a user's preference information acquired in advance. The voice recognition method is characterized in that the recognition result is determined by using the preference score of the recognition target vocabulary, and by adding the preference score to the recognition score, the user is more likely to speak. It has the effect of weighting the content and outputting the result, greatly reducing the cases where the user's completely unintended result appears, and at the same time, causing the intended result to appear with a high probability.

[0011]
The invention according to claim 3 is the voice recognition method according to claim 1 or 2 , wherein the user's preference information selects a user identification symbol, a user's voice, a user's image, or the user's preference information input by the user. It is a voice recognition method that is at least one piece of information selected from the time to be performed, and is related to learning the preference information of the present invention. It has the effect of being able to learn tastes without spending time. Furthermore, by using learning using likes or dislikes, it is possible to learn preference information without impairing the promptness of voice operation and the ease of handling.

[0012]
The invention according to claim 4 is a program designation method for designating a program by using the voice recognition method according to any one of claims 1 to 3, and is used for digital television broadcasting and on-de-mand movies. / It is applied to the use of the karaoke distribution service, and has the effect of being able to select a program that suits the user's taste.
The invention according to claim 5 is the program designation method according to claim 4, wherein the user's preference information is information to be learned when the program is interrupted by the viewing side, when the program is interrupted by the distribution side, or when the program ends. By learning the user's preference information after viewing the program content, it has the effect of being able to select a program that better suits the user's preference.
The invention according to claim 6 is the program designation method according to claim 4 or 5, wherein the user's preference information is information learned by using the viewing history of the program, and the viewing history of the program is used. Since the user's preference information is learned, the user's trouble is saved and the user's preference can be easily adjusted.
It has the effect of being able to learn properly.
According to the invention of claim 7, the viewing history of the program is the time zone, genre, performer, appearance group name, program name, program content, theme, music, content keyword, and user name provided by the electronic program guide. The program designation method according to claim 6, which is information having at least one as a component, is used, and by designating the content of the user's preference information, the program can be selected more according to the user's preference. Has.
The invention according to claim 8 is to obtain a recognition target vocabulary from a voice input unit for inputting a voice uttered by a user and a recognition target candidate vocabulary storing all words to be voice recognition based on the user's preference information. It is a voice recognition device including a recognition target vocabulary creation unit to be created, a recognition unit that recognizes a word from the voice input from the voice input unit and the recognition target vocabulary, and recognizes according to the user's preference. It has the effect of improving speech recognition performance by narrowing down the target vocabulary.

0013
The invention according to claim 9 creates a recognition target vocabulary from a voice input unit for inputting a voice uttered by the user and a recognition target candidate vocabulary storing all words to be voice recognition based on the user's preference information. A recognition target vocabulary creation unit, a preference score calculation unit that calculates a preference score of the recognition target vocabulary based on the user's preference information, a voice input from the voice input unit, and a word from the recognition target vocabulary. It is a voice recognition device including a recognition unit for calculating a recognition score and a recognition result determination unit for determining and outputting a recognition result using the preference score and the recognition score, according to the user's preference. It has the effect of improving speech recognition performance by narrowing down the vocabulary to be recognized.

0014.
The invention according to claim 10 further includes a preference information storage unit that holds preference information of one or more users, and the user preference information includes a user identification symbol input by the user, a user's voice, and a user. The voice recognition device according to claim 8 or 9, which is information selected by at least one of the time to select the image or the user's preference information, and is a recognition target vocabulary from one or more preference information. By selecting the preference information used for selection, it has the effect of improving the speech recognition performance by narrowing down the recognition target vocabulary according to the user's preference.

[0019]
The invention according to claim 11 is a program designation device characterized in that a program is designated by using the voice recognition device according to any one of claims 8 to 10, and is used for digital television broadcasting and on. -It is applied to the program designation device when using the de Mand movie / karaoke distribution service, and has the effect of being able to select a program that suits the user 's taste.

0020
The program according to claim 12 , further comprising a preference information creating unit that prompts the user to input preference information at the time of interruption by the viewing side of the program, interruption by the distribution side of the program, or the end of the program. It is a designated device, and when the program designation device learns preference information from the user, it is possible to select a program that better suits the user's preference by actively asking for the preference, aiming at the break of the program. It has the effect of.
The invention according to claim 13 is the program designation device according to claim 11 or 12, wherein the user's preference information is information learned by using the viewing history of the program, and the preference is obtained by using the viewing history of the program. Since information is learned, it saves the user 's time and effort and has the effect of easily learning according to the user's taste.

0021.
According to the invention of claim 14 , the viewing history of the program is the time zone, genre, performer , appearance group name, program name, program content, theme, music, content keyword, user name provided by the electronic program guide. The program designation device according to claim 13, which is information having at least one of the above as a component, and by designating the content of the preference information, it is possible to select a program more suitable for the user 's preference. It has the effect of.

[0022]
The invention according to claim 15 creates a recognition target vocabulary based on user preference information acquired in advance from a recognition target candidate vocabulary storing all words to be voice recognition, and recognizes the input voice. A program for causing a computer to execute a voice recognition method characterized by recognizing a word from a target vocabulary. It has an effect of improving voice recognition performance by narrowing down the recognition target vocabulary according to the user's preference. In addition, it is a method that is simpler and faster than performing speech recognition for all vocabularies that can be recognition target candidates.
The invention according to claim 16 is based on a recognition score obtained by voice recognition for recognizing the word from a input voice and a recognition target vocabulary which is a word to be voice-recognized, and a user's preference information acquired in advance. It is a program for causing a computer to execute a voice recognition method characterized by determining a recognition result by using the preference score of the recognition target vocabulary, and the preference score is also added to the recognition score. As a result, the content that is more likely to be spoken by the user is weighted and the result is output, which greatly reduces the cases where the user's completely unintended result appears, and at the same time, has the effect of causing the intended result to appear with a high probability. ..

[Explanation of code]
1 Voice input unit 2 Preference information creation unit 3 Preference information storage unit 4 Recognition target candidate vocabulary storage unit 5 Recognition target vocabulary creation unit 6 Recognition unit 7 Recognition result determination unit 8 Set top box 9 Digital TV broadcast 10 On-de-mand movie / Karaoke distribution service 11 Video 12 Display monitor 13 Sound template storage unit 14 Sound template selection unit 15 Preference selection information input unit 16 Number determination unit 17 Preference score calculation unit

Claims (16)

Create in advance a recognition target vocabulary based on user preference information acquired in advance from recognition target candidate vocabularies in which all words to be subjected to speech recognition are stored, and recognize words from the input speech and the recognition target vocabulary A speech recognition method characterized by A recognition score obtained by speech recognition for recognizing the word from the input speech and a recognition target vocabulary which is a word for speech recognition, and a preference score of the recognition target vocabulary based on preference information of the user acquired in advance A speech recognition method, characterized in that the recognition result is determined using. The voice according to claim 1 or 2, wherein the preference information of the user is at least one information selected from a user identification symbol input by the user, an utterance of the user, an image of the user or a time of selecting the preference information of the user. Recognition method. A program specification method for specifying a program using the speech recognition method according to any one of claims 1 to 3. 5. The program specification method according to claim 4, wherein the preference information of the user is information to be learned at the time of interruption by the program viewing side, at the time of interruption by the program distribution side or at the end of the program. 6. The program specification method according to claim 4, wherein the preference information of the user is information learned using a program viewing history. The viewing history of the program is information having at least one of the time zone, genre, cast, cast group name, program name, program content, theme, music, content keyword, and user name provided by the electronic program guide as a component The program designation method according to claim 6, which is A voice input unit for inputting a voice uttered by a user; a recognition target vocabulary creation unit for creating a recognition target vocabulary from recognition target candidate vocabulary storing all words to be voice recognition targets based on user preference information; A speech recognition apparatus comprising: a recognition unit that recognizes a word from the speech input from the speech input unit and the recognition target vocabulary. A voice input unit for inputting a voice uttered by a user; a recognition target vocabulary creation unit for creating a recognition target vocabulary from recognition target candidate vocabulary storing all words to be voice recognition targets based on user preference information; recognition of calculating a preference score calculation unit for calculating a preference score of the recognition target words based on the preference information of the user, the recognition score of a word and a speech before Ki認 identification target words that are input from the voice input unit parts and uses and the preference scores and before Ki認 identification scores, the speech recognition system comprising a recognition result determining unit that determines a recognition result output. The information processing apparatus further includes a preference information storage unit holding preference information of one or more users, wherein the user preference information is a user identification symbol input by the user, the user
The speech recognition apparatus according to claim 8 or 9, wherein the information is selected by at least one of an utterance of-, an image of a user, and a time of selecting preference information of the user . Using the speech recognition apparatus of any serial mounting of claims 8 to 10, the program designation apparatus characterized by the program designation. Time of the interruption by the program viewing聴側, when interruption or when the program ends by the distribution side of the program, further comprising <br/> claim 11 program designation device according to the preference information creating unit that prompts the input of the user preference information. 13. The program specification device according to claim 11, wherein the preference information of the user is information learned using a program viewing history. Program viewing history has between band when it is provided by the electronic program guide, genre, cast, starring the group name, program name, program content, theme, music, content keywords, as at least one of the components of a user name The program specification device according to claim 13, which is information . Create in advance a recognition target vocabulary based on user preference information acquired in advance from recognition target candidate vocabularies in which all words to be subjected to speech recognition are stored, and recognize words from the input speech and the recognition target vocabulary A program for causing a computer to execute a speech recognition method characterized by A recognition score obtained by speech recognition for recognizing the word from the input speech and a recognition target vocabulary which is a word for speech recognition, and a preference score of the recognition target vocabulary based on preference information of the user acquired in advance A program for causing a computer to execute a speech recognition method, characterized by determining a recognition result using.