Voice-activated interactive speech recognition device and method

• US 5,983,186 A
• Filed: 08/20/1996
• Issued: 11/09/1999
• Est. Priority Date: 08/21/1995
• Status: Expired due to Term

• Current Assignee: Seiko Epson Corporation (Seiko Group)
• Original Assignee: Seiko Epson Corporation (Seiko Group)
• Inventors: Miyazawa, Yasunaga, Inazumi, Mitsuhiro, Hasegawa, Hiroshi, Urano, Osamu, Edatsune, Isao
• Primary Examiner(s): Hudspeth, David R.
• Assistant Examiner(s): ABEBE, DANIEL DEMELASH
• Application Number:
  US08/700,181
• Time in Patent Office: 1,176 Days
• Field of Search: 179/15, 179/251, 704/212, 704/275, 704/246, 704/233, 704/255, 381/57
• US Class Current: 704/275
• CPC Class Codes: G10L 15/26 Speech to text systems G10L15/08 takes precedence
• CPC Class Codes: G10L 2025/783 based on threshold decision

Abstract

Techniques for implementing adaptable voice activation operations for interactive speech recognition devices and instruments. Specifically, such speech recognition devices and instruments include an input sound signal power or volume detector in communication with a central CPU for bringing the CPU out of an initial sleep state upon detection of perceived voice exceeding a predetermined threshold volume level and is continuously perceived for at least a certain period of time. If both these conditions are satisfied, the CPU is transitioned into an active mode so that the perceived voice can be analyzed against a set of registered key words to determine if a "power on" command or similar instruction has been received. If so, the CPU maintains an active state in normal speech recognition processing ensues until a "power off" command is received. However, if the perceived and analyzed voice can not be recognized, it is deemed to be background noise and the minimum threshold is selectively updated to accommodate the volume level of the perceived but unrecognized voice. Other aspects include tailoring the volume level of the synthesized voice response according to the perceived volume level as detected by the input sound signal power detector, as well as modifying audible response volume in accordance with updated volume threshold levels.

Claims

1. A method for implementing transient, voice-based activation in CPU-driven speech recognition equipment including a speech recognition CPU and a low power consuming input sound signal power detector, comprising the steps of:

(A) placing the speech recognition CPU in a sleep mode wherein the speech recognition CPU is inactive but the input sound signal power detector is active;

(B) detecting whether a first sound continuously exceeds a predetermined volume for at least a predetermined time period;

(C) enabling the speech recognition CPU to determine whether the detected first sound contains at least one keyword recognized by the speech recognition CPU as an activation command, and performing one of the following steps, if it is detected that the first sound continuously exceeds the predetermined volume for at least the predetermined time period;

(1) activating the speech recognition CPU, if the detected first sound contains the at least one keyword recognized by the speech recognition CPU as an activation command, or

(2) returning the speech recognition CPU to the sleep mode if the detected first sound fails to contain the at least one keyword recognized by the speech recognition CPU as an activation command.

2. The method of claim 1, further comprising the steps of:

determining whether a received second sound contains at least one keyword recognized by the speech recognition CPU as a deactivation command, if the detected first sound contains the at least one keyword recognized by the speech recognition CPU as an activation command; and

returning the speech recognition CPU to the sleep mode, if the detected second sound contains the at least one keyword recognized by the speech recognition CPU as a deactivation command.

3. The method of claim 1, wherein the first sound has first volume level greater than the threshold volume, and wherein step (C) further comprises issuing an audible response at a second volume level substantially equal to the first volume level.

4. The method of claim 1, wherein, in step (C)(2), the threshold volume is selectively increased if the detected first sound fails to contain the at least one keyword recognized by the speech recognition CPU as an activation command.

5. A voice-activated speech recognition apparatus, comprising:

a sound signal input unit that receives a sound and translates the received sound into digital form;

a speech recognition CPU having an active mode and a sleep mode, comprising;

a sound signal analyzer in communication with said sound signal input unit to generate voice feature parameters for the received digitized input sound, when said speech recognition CPU is in the active mode;

a phrase detector in communication with said sound signal analyzer to determine whether the feature parameters substantially match speech reference template information corresponding to pre-registered expressions and generate phrase detection data in response thereto, when said speech recognition CPU is in the active mode; and

a speech comprehension interaction controller in communication with said phrase detector to recognize a meaning from the received digitized input sound based on the received phrase detection data and conversation content and to formulate an appropriate response corresponding to the recognized meaning, when said speech recognition CPU is in the active mode; and

a low power consuming input sound signal power detector in communication with said sound signal input unit and said speech recognition CPU, wherein, when said speech recognition CPU is in the sleep mode, said input sound signal power detector remains active and determines whether the received input sound exceeds a threshold volume for at least a predetermined time period, and wherein said input signal power detector enables said speech recognition CPU, if the received input sound exceeds the threshold volume for at least the predetermined time period, to determine whether the received input sound contains at least one keyword recognized by said speech recognition CPU as an activation command, and cooperates in placing said speech recognition CPU in the active mode if said speech recognition CPU determines that the received input sound contains the at least one keyword that corresponds to an activation command.

6. The apparatus of claim 5, wherein said input sound signal power detector selectively increases the threshold volume if the received input sound exceeds the threshold volume for at least the predetermined time period but said speech recognition CPU determines the at least one input sound does not contain the at least one keyword that corresponds to an activation command.

7. The apparatus of claim 5, wherein the input signal power detector comprises:

a first determining circuit for determining whether the received input sound exceeds the threshold volume for at least the predetermined time period;

a communication mechanism responsive to said first determining circuit and communicating with said speech recognition CPU for placing the speech recognition CPU into the active mode, if the received input sound exceeds the threshold volume for at least the predetermined time period and said speech recognition CPU determines that the received input sound contains the at least one keyword that corresponds to an activation command; and

a threshold volume adjuster in communication with said first determining circuit and said speech recognition CPU for selectively increasing the threshold volume, if the received input sound exceeds the threshold volume for the predetermined time period but said speech recognition CPU determines that the received input sound does not contain the at least one keyword that corresponds to an activation command.

8. The apparatus of claim 5, wherein the speech recognition CPU further comprises a speech synthesizer in communication with said speech comprehension interaction controller to generate synthesized audio corresponding to the appropriate response formulated by said speech comprehension interaction controller, when the speech recognition CPU is in the active mode.

9. The apparatus of claim 8, wherein said speech recognition CPU further comprises a response volume adjuster responsive to said input signal power level detector and said speech synthesizer for tailoring the volume level of the appropriate response based on the perceived volume level of the given input sound signal.