System for registering speech information to make a voice dictionary

• US 4,831,653 A
• Filed: 11/16/1987
• Issued: 05/16/1989
• Est. Priority Date: 11/12/1980
• Status: Expired due to Term

• Current Assignee: Canon Kabushiki Kaisha (Canon Inc.)
• Original Assignee: Canon Kabushiki Kaisha (Canon Inc.)
• Inventors: Katayama, Hirohiko
• Primary Examiner(s): KEMENY, EMANUEL
• Application Number:
  US07/120,788
• Time in Patent Office: 547 Days
• Field of Search: 381/29-53, 364/513, 364/513.5
• US Class Current: 704/243
• CPC Class Codes: G10L 15/063 Training
• CPC Class Codes: G10L 15/22 Procedures used during a speech recognition process, e.g. man-machine dialogue

Abstract

A system for having a mode for registering storing a necessary typical spoken word pattern for use in speech recognition and inputting a same spoken word a plurality of times. In the mode for registering a word pattern, a previous input spoken word, input prior to the Nth time, is reproduced. After listening to the reproduced spoken prompt word, the Nth input of the same spoken word is effected. The system registers one typical spoken word of a well-averaged voice pattern with high accuracy while increasing the system speech recognition rate.

Claims

1. A system for registering speech information as a typical pattern for use in a word recognition system comprising:

input means for inputting spoken words;

mode change means for changing between a word-registration mode and a word-recognition mode;

memory means in said word-registration mode for storing a spoken word obtained on the basis of inputting the same spoken word a plurality of times for use in said word-recognition mode; and

output means for reading, when said system is in the registration mode and selected spoken word is input for the Nth time, said selected spoken word stored in said memory means before the Nth time, and said output means for outputting said selected spoken word as spoken information.

2. The system for claim 1, wherein said memory means comprises a first memory means for storing averaged speech pattern information and a second memory means for storing the spoken information.

3. The system of claim 2, wherein the averaged speech pattern to be stored in the first memory means and an address specifying a storage location of spoken information to be stored in the second memory means in response to the averaged speech pattern are associated with each other.

4. The system of claim 1, wherein said input means effects the input of said selected spoken word for the Nth time after output of said selected spoken word from said output means.

5. A system for registering speech information as a typical pattern for use in a word recognition system comprising:

input means for inputting a spoken word from an operator a plurality of times;

mode change means for changing between a word-registration mode and a word-recognition mode;

memory means for storing a spoken word pattern input by said input means for use in a word-recognition mode, said word-registration mode storing a typical spoken word pattern for recognition in said memory means on the basis of inputting said spoken word a plurality of times; and

output means for reading said spoken word stored in said memory before the Nth time and for outputting said stored word as spoken information upon the input of said spoken word by said operator for the Nth time when said system is in the registration mode,

wherein said system further comprises designation means for designating a word pattern registered in said memory means and said output means reads said spoken word in order to re-register the word pattern designated by said designation means.

6. The system of claim 5, wherein said mode change means terminates the registration mode when said operator repeats the input of said spoken word for the Nth time after listening to the spoken information output by said output means in response to the input of the spoken word for the (N-1)th time when said system is in the registration mode.