Speech recognition method, and apparatus and computer controlled apparatus therefor

A recognition method for recognizing speech data, comprising:

• a description step of describing a state transition model for the unknown word, represented by the number of the states and the transition probability among the states;
• an acquisition step of acquiring a state of the maximum likelihood at each time of the speech data, among the plural states employed in the state transition model for the known word, to be employed in the speech recognition of the known word;
• an output step of acquiring the likelihood of a state transition model, obtained by applying the state acquired by said acquisition step to said state transition model for the unknown word, to said speech data, and outputting said likelihood as the transition model of the unknown word and the likelihood therefor;
• and
• a recognition step of effecting recognition of said speech data, based on the likelihood obtained with said state transition model for the known word and on the state transition model and the likelihood for the unknown word.

A computer controlled apparatus for effecting a speech recognition process by reading a predetermined program from a memory medium, wherein said memory medium comprises:

• a state transition model for the unknown word, represented by the number of the states and the transition probability among the states;
• a process code for an acquisition step of acquiring a state of the states maximum likelihood at each time of the speech data, among the plural states employed in the state transition model for the known word, to be employed in the speech recognition of the known word;
• a process code for an output step of acquiring the likelihood of a state transition model, obtained by applying the state acquired by said acquisition step to said state transition model for the unknown word, to said speech data, and outputting said likelihood as the transition model of the unknown word and the likelihood therefor;
• and
• a process code for a recognition step of effecting recognition of said speech data, based on the likelihood obtained with said state transition model for the known word and on the state transition model and the likelihood for the unknown word.