Device control device and device control method

• US 7,505,909 B2
• Filed: 12/06/2004
• Issued: 03/17/2009
• Est. Priority Date: 12/05/2003
• Status: Active Grant

• Current Assignee: Rakuten Group, Inc.
• Original Assignee: Kabushiki Kaisha Kenwood (JVC Kenwood Corporation)
• Inventors: Koyama, Rika
• Primary Examiner(s): Chawan; Vijay B
• Application Number:
  US10/581,822
• Publication Number:
  US 20070239443A1
• Time in Patent Office: 1,562 Days
• Field of Search: 704/231, 704/275, 704/246, 704/270, 704/240, 340/825.72, 340/5.72, 340/5.1
• US Class Current: 704/275
• CPC Class Codes: G10L 15/22 Procedures used during a speech recognition process, e.g. man-machine dialogue
• CPC Class Codes: G10L 2015/228 of application context

Abstract

A language analyzer (2) performs speech recognition on a speech input by a speech input unit (1), specifies a possible word which is represented by the speech, and the score thereof, and supplies word data representing them to an agent processing unit (6). The agent processing unit (6) stores process item data which defines a data acquisition process to acquire word data or the like, a discrimination process, and an input/output process, and wires or data defining transition from one process to another and giving a weighting factor to the transition, and executes a flow represented generally by the process item data and the wires to thereby control devices belonging to an input/output target device group (5). To which process in the flow the transition takes place is determined by the weighting factor of each wire, which is determined by the connection relationship between a point where the process has proceeded and the wire, and the score of word data. The wire and the process item data can be downloaded from an external server.

Claims

1. A device control device for controlling a plurality of devices, each being controlled by use of numerical parameters, the device control device comprising:

constant storage means that stores in advance a phrase and a predetermined constant in association with each other;

speech recognition means that acquires speech data representing a speech, and performs speech recognition on said speech data, thereby specifying a candidate for a phrase included by said speech and computes a likelihood for each specified candidate; and

device control means that specifies those devices which are controllable of said plurality of devices, and an amount of change of said numerical parameter based on the predetermined constant associated with the specified phrase and the likelihood that have been computed by said speech recognition means, and changes the numerical parameter by the specified amount of change to thereby control the specified devices, wherein

said device control means controls a device when a number of said specified devices is one, and

when there are a plurality of devices specified, said speech recognition means further acquires speech data to specify a candidate for the phrase and computes the likelihood, and

said device control means specifies any one of the devices that are controllable, based on the predetermined constant associated with the specified phrase and likelihood computed by said speech recognition means, and controls said specified device.

2. The device control device according to claim 1, wherein when there are a plurality of devices specified, said device control means outputs data prompting a user to utter a speech that specifies any one of said controllable devices.

3. The device control device according to claim 1, wherein

said candidate for the phrase specified by said speech recognition means represents a meaning of increasing or decreasing said numerical parameter;

one of said plurality of devices is an audio device, and a numerical parameter for controlling said audio device is a sound volume; and

another device is a power window, and a numerical parameter for controlling said power window is an opening/closing amount of a window.

4. The device control device according to claim 1, wherein

said candidate for the phrase specified by said speech recognition means represents a meaning of increasing or decreasing said numerical parameter;

one of said plurality of devices is an air conditioner, and a numerical parameter for controlling said air conditioner is a temperature, and

another device is a power window, and a numerical parameter for controlling said power window is an opening/closing amount of a window.

5. The device control device according to claim 1, wherein

said candidate for the phrase specified by said speech recognition means represents a meaning of increasing or decreasing said numerical parameter;

one of said plurality of devices is an air conditioner, and a numerical parameter for controlling said air conditioner is a temperature, and

another device is an audio device, and a numerical parameter for controlling said audio device is a volume.

6. The device control device according to claim 1, wherein

said device control means computes a value represented by the predetermined constant multiplied by said likelihood, and specifies said controllable device and the amount of change of said numerical parameter, based on a candidate for a phrase that corresponds to the largest of the values.

7. A device control method for controlling a plurality of devices, each being controlled by use of numerical parameters, comprising:

a speech recognition step of acquiring speech data representing a speech, and performing speech recognition on said speech data, thereby specifying a candidate for a phrase included by said speech and computing a likelihood for each specified candidate; and

a device control step of specifying those devices which are controllable of said plurality of devices, and an amount of change of said numerical parameter based on the predetermined constant associated with the specified phrase and the likelihood that have been computed in said speech recognition step, and changing the numerical parameter by the specified amount of change to thereby control the specified devices, wherein

in said device control step,

a device is controlled when a number of specified devices is one, and

when there are a plurality of devices specified, said speech recognition step further acquires speech data to specify a candidate for the phrase and computes the likelihood, and

said device control step specifies any one of the devices that are controllable, based on the predetermined constant associated with the specified phrase and the likelihood computed by said speech recognition means, and control said specified device.

8. A computer-readable recording medium storing a program which allows a computer that controls a plurality of devices by use of a numerical parameter, to execute:

a speech recognition step of acquiring speech data representing a speech, performing speech recognition on said speech data, thereby specifying a candidate for a phrase included by said speech and computing a likelihood for each specified candidate; and

a device control step of specifying those devices which are controllable of said plurality of devices, and an amount of change of said numerical parameter based on the predetermined constant associated with the specified phrase and the likelihood that have been computed by said speech recognition step, and changing the numerical parameter by the specified amount of change to thereby control the specified devices, wherein

said device control step controls a device when a number of specified devices is one, and

when there are a plurality of devices specified, said speech recognition step further acquires speech data to specify a candidate for the phrase and computes the likelihood, and

said device control step specifies any one of the devices that are controllable, based on the predetermined constant associated with the specified phrase and the likelihood computed by said speech recognition means, and controls said specified device.