Speech recognition system for sequence data

A speech recognition system for recognizing voices of sequence data which ought to be inputted in regular, predetermined order, comprising:

• a dummied-sequence-data storage unit storing data which have been obtained by adding at least one item of dummy data in all items of the sequence data which ought to be inputted in regular, predetermined order to each item of the sequence data as ought to be currently inputted in predetermined order;
• a speech recognition unit recognizing and reading out the sequence data item which ought to be currently inputted in the predetermined order, as well as at least one dummy data item which corresponds to said sequence data item, from said dummied-sequence data storage unit, thereby to form a set of recognition reference words for the speech recognition, and outputting the recognition reference word which is the most similar to an externally inputted voice among the recognition reference words, as a recognized result on the basis of a feature of the input voice;
• and
• a decision unit deciding by comparing the recognized result of said speech recognition unit with the data item which ought to be currently inputted in the predetermined order, so as to decide a match between said recognized result and said data item.

A computer-readable storage medium storing a program for recognizing spoken words based upon and corresponding to sequence data which ought to be inputted in regular, predetermined order, said program causing a computer to execute:

• generating a recognition reference word by adding at least one item of dummy data for all items of the sequence data which ought to be inputted in regular, predetermined order to each item of the sequence data as ought to be currently inputted in predetermined order, thereby to form a set of recognition reference words for speech recognition;
• comparing an inputted word included in the spoken words with each word in the set of the recognition reference words;
• determining whether characteristics of the inputted word are most similar to those of one of the words included in the set of recognition reference words as a word to be inputted;
• recognizing and outputting the recognition reference word which is the most similar to the externally inputted word among the recognition reference words, as a recognized result on the basis of a feature of the input spoken word;
• and
• comparing the recognized result with the each item of the sequence data as ought to be currently inputted in the predetermined order, so as to decide a match between said recognized result and said each item of the sequence data as ought to be currently inputted in the predetermined order.

A computer-readable storage medium storing a program for recognizing spoken words based upon corresponding to sequence data which ought to be inputted in regular, predetermined order, said program causing a computer to execute:

• reading out the sequence data item from a dummied-sequence-data storage unit which stores data obtained by adding at least one item of dummy data in all items of the sequence data which ought to be inputted in regular, predetermined order to each sequence data item, the sequence data item which is to be currently inputted in predetermined order, as well as at least one dummy data item which corresponds to said sequence data item, thereby to form a set of recognition reference words for speech recognition;
• comparing an inputted word of the spoken words with each word in the set of the recognition reference words;
• determining whether characteristics of the inputted word are most similar to those of a word included in the set of the recognition reference words as a word to be inputted;
• and
• outputting the recognition reference word which is the most similar to the externally inputted word among the recognition reference words, as a recognized result on the basis of a feature of the input spoken word;
• and
• comparing the recognized result with the each item of the sequence data as ought to be currently inputted in the predetermined order, so as to decide a match between said recognized result and said each item of the sequence data as ought to be currently inputted in the predetermined order.

A speech recognition method for recognizing spoken words corresponding to sequence data which ought to be inputted in regular, predetermined order, comprising:

• generating a recognition reference word by adding at least one item of dummy data for all items of the sequence data which ought to be inputted in regular, predetermined order to each item of the sequence data as ought to be currently inputted in predetermined order, thereby to form a set of recognition reference words for speech recognition;
• comparing an inputted word included in the spoken words with each word in the set of the recognition reference words;
• determining whether characteristics of the inputted word are most similar to those of one of the words included in the set of recognition reference words as a word to be inputted;
• recognizing and outputting the recognition reference word which is the most similar to the externally inputted word among the recognition reference words, as a recognized result on the basis of a feature of the input spoken word;
• and
• comparing the recognized result with the each item of the sequence data as ought to be currently inputted in the predetermined order, so as to decide a match between said recognized result and said each item of the sequence data as ought to be currently inputted in the predetermined order.

A speech recognition method for recognizing spoken words corresponding to sequence data which ought to be inputted in regular, predetermined order, comprising:

• reading out the sequence data item from a dummied-sequence-data storage unit which stores data obtained by adding at least one item of dummy data in all items of the sequence data which ought to be inputted in regular, predetermined order to each sequence data item, the sequence data item which is to be currently inputted in predetermined order, as well as at least one dummy data item which corresponds to said sequence data item, thereby to form a set of recognition reference words for speech recognition;
• comparing an inputted word of the spoken words with each word in the set of the recognition reference words;
• determining whether characteristics of the inputted word are most similar to those of a word included in the set of the recognition reference words as a word to be inputted;
• outputting the recognition reference word which is the most similar to the externally inputted word among the recognition reference words, as a recognized result on the basis of a feature of the input spoken word;
• and
• comparing the recognized result with the each item of the sequence data as ought to be currently inputted in the predetermined order, so as to decide a match between said recognized result and said each item of the sequence data as ought to be currently inputted in the predetermined order.