Voice recognition system using voice power patterns

• US 4,827,519 A
• Filed: 09/17/1986
• Issued: 05/02/1989
• Est. Priority Date: 09/19/1985
• Status: Expired due to Fees

• Current Assignee: Ricoh Company Limited
• Original Assignee: Ricoh Company Limited
• Inventors: Muroi, Tetsuya, Fujimoto, Junichiroh
• Primary Examiner(s): Roskoski, Bernard
• Assistant Examiner(s): Hoff, Marc S.
• Application Number:
  US06/908,681
• Time in Patent Office: 958 Days
• Field of Search: 381/41, 381/42, 381/43, 381/45
• US Class Current: 704/250
• CPC Class Codes: G10L 15/02 Feature extraction for speech recognition; Selection of recognition unit

Abstract

A voice processing method and system using voice power is provided. A voice power signal is produced in analog format and this voice power signal is sampled at a predetermined time interval to determine its current envelope or amplitude, which is then compared with a predetermined number of voice power levels, whereby "1" is assigned to at least one of the predetermined number of voice power levels which corresponds in level to the amplitude thus determined and "0" is assigned to the rest of the predetermined number of voice power levels, thereby converting the analog voice power signal into a binary voice power pattern. In the preferred embodiment, the voice signal is also processed through a frequency analyzer including a plurality of band-pass filters different in frequency range and a binary converter to form a binary time-frequency voice distribution pattern which is then combined with the binary voice power pattern to form a combined voice pattern. As a modification, a voice power range is determined by multiplying a predetermined threshold to the amplitude, and the voice power pattern is compared with the predetermined number of voice power levels.

Claims

1. A method for forming a binary voice power pattern, comprising the steps of:

converting a voice in the form of pressure waves into an electrical voice signal;

processing said electrical voice signal to define an analog voice power signal whose voice power is equal to or above a predetermined voice power level;

sampling said analog voice power signal at a predetermined time interval to determine a current amplitude thereof; and

comparing the thus determined current amplitude with a predetermined number of voice power levels, thereby assigning a first binary data to at least one of the predetermined number of voice power levels which corresponding to said amplitude and assigning a second binary data, which is different from said first binary data, to the remaining of the predetermined number of voice power levels to form said binary voice power pattern.

2. The method of claim 1 further comprising, after the step of sampling, a step of determining a voice power range by multiplying a predetermined threshold to said amplitude, and wherein said step of comparing compares said voice power range with said predetermined number of voice power levels, thereby assigning said first binary data to at least one of the predetermined number of voice power levels which corresponds to said voice power range and said second binary data to the remaining of the predetermined number of voice power levels to form said binary voice power pattern.

3. A voice processing system comprising:

a transducer for converting a voice in the form of pressure waves into an electrical voice signal;

processing means for processing said electrical voice signal to define an analog voice power signal whose voice power is equal to or higher than a predetermined voice power level;

sampling means for sampling said analog voice power signal at a predetermined time interval to determine an amplitude thereof;

comparing means for comparing said amplitude with a predetermined number of voice power levels to assign a first binary data to at least one of said predetermined number of voice power levels which corresponds to said amplitude and a second binary data, which is different from said first binary data, to the remaining of said predetermined number of voice power levels, thereby converting said analog voice power signal into a binary voice power pattern; and

storing means for storing said binary voice power pattern.

4. The system of claim 3 further comprising adding means for adding at least one other binary voice power pattern formed by said sampling means to said binary voice power pattern stored in said storing means in a superposed relationship.

5. The system of claim 3 further comprising comparing means for comparing said binary voice pattern formed by said sampling means with said binary voice power pattern stored in said storing means, and means for determining a degree of similarity between said two compared patterns.

6. A voice processing system comprising:

a transducer for converting a voice in the form of pressure waves into an electrical voice signal;

processing means for processing said electrical voice signal to define an analog voice signal whose voice power is equal to or higher than a predetermined voice power level;

first detecting means for detecting a voice power of said analog voice signal to form an analog voice power signal;

second detecting means for detecting a predetermined feature other than voice power of said analog voice signal to form at least one voice feature signal;

sampling means for sampling said analog voice power signal at a predetermined time interval to determine an amplitude thereof;

comparing means for comparing said amplitude with a predetermined number of voice power levels to assign a first binary data to at least one of said predetermined number of voice power levels which corresponds to said amplitude and a second binary data, which is different from said first binary data, to the remaining of said predetermined number of voice power levels, thereby converting said analog voice power signal into a binary voice power pattern, said sampling means also sampling said voice feature signal and converting said voice feature signal into a voice feature binary pattern in a predetermined manner, which is then combined with said binary voice power pattern to define a combined binary voice pattern; and

storing means for storing said combined binary voice power pattern.

7. The system of claim 6 wherein said second detecting means includes a predetermined number of band-pass filters different im frequency, so that said predetermined feature is a time-frequency distribution pattern of said voice.

8. The system of claim 7 further comprising adding means for adding at least one other combined binary voice pattern formed by said sampling means to said combined binary voice pattern stored in said storing means as superposed thereon.

9. The system of claim 8 further comprising comparing means for comparing said combined binary voice pattern formed by said sampling means with said combined binary voice pattern stored in said storing means and means for determining a degree of similarity between said two compared patterns.

10. The system of claim 9 wherein said comparing means includes partial comparing means for comparing said two pattern partially, preliminary calculating means for calculating a preliminary degree of similarity between said partially compared patterns and judging means for determining whether or not said two compared patterns are to be fully compared to calculate the degree of similarity between the two patterns based on said calculated preliminary degree of similarity.

11. A system of processing voice information comprising:

transducer means for converting a voice in the form of pressure waves into an electrical voice signal;

processing means for processing said electrical voice signal to define an analog voice power signal whose voice power is equal to or higher than a predetermined voice power level;

means for sampling said analog voice power signal at a predetermined time interval to determine an amplitude thereof;

means for determining a voice power range by multiplying a predetermined threshold to said amplitude;

means for comparing said voice power range with a predetermined number of voice power levels;

means for assigning a first binary data to at least one of said predetermined number of voice power levels which corresponds to said voice power range and a second binary data, which is different from said first binary data, to the remaining of said predetermined number of voice power levels, thereby converting said analog voice power signal into a binary voice power pattern; and

storing means for storing said binary voice power pattern.