Command recognition device, command recognition method, and command recognition robot

• US 8,532,989 B2
• Filed: 09/02/2010
• Issued: 09/10/2013
• Est. Priority Date: 09/03/2009
• Status: Active Grant

• Current Assignee: Honda Motor Co., Ltd. (Honda Motor Company)
• Original Assignee: Honda Motor Co., Ltd. (Honda Motor Company)
• Inventors: Funakoshi, Kotaro, Nakano, Mikio, Zuo, Xiang, Iwahashi, Naoto, Taguchi, Ryo
• Primary Examiner(s): COLUCCI, MICHAEL C
• Application Number:
  US12/874,779
• Publication Number:
  US 20110112839A1
• Time in Patent Office: 1,104 Days
• Field of Search: 704/275, 704/235, 704/231, 704/10, 704/2, 704/215, 704/249, 704/254, 704/257, 348/36, 700/245, 700/250, 701/25, 706/23, 706/45, 715/716
• US Class Current: 704/240
• CPC Class Codes: G10L 15/083 Recognition networks G10L15/142, G10L15/16 take precedence
• CPC Class Codes: G10L 15/197 Probabilistic grammars, e.g. word n-grams

Abstract

A command recognition device includes: an utterance understanding unit that determines or selects word sequence information from speech information; speech confidence degree calculating unit that calculates degree of speech confidence based on the speech information and the word sequence information; a phrase confidence degree calculating unit that calculates a degree of phrase confidence based on image information and phrase information included in the word sequence information; and a motion control instructing unit that determines whether a command of the word sequence information should be executed based on the degree of speech confidence and the degree of phrase confidence.

Claims

1. A command recognition device comprising:

an utterance understanding unit that determines or selects word sequence information from speech information;

speech confidence degree calculating unit that calculates degree of speech confidence based on the speech information and the word sequence information;

a phrase confidence degree calculating unit that calculates a degree of phrase confidence based on image information and phrase information included in the word sequence information;

an image analysis unit that calculates a trajectory-of-motion of an object from the image information, wherein the trajectory of motion is calculated based on time-series data of position coordinates of a trajectory in a motion of an object, where the position coordinates are variable in time; and

a motion control instructing unit that determines whether a command of the word sequence information is to be executed based on a command estimation value which is calculated using the degree of speech confidence and the degree of phrase confidence,

wherein the phrase confidence degree calculating unit calculates a degree of motion confidence representing the confidence that the trajectory-of-motion comprises a motion of the phrase information as the degree of phrase confidence.

2. The command recognition device according to claim 1, wherein the image analysis unit calculates a feature quantity of an object from the image information, and

wherein the phrase confidence degree calculating unit calculates an degree of image confidence representing the confidence that the object having the feature quantity is an object of the phrase information as the degree of phrase confidence.

3. The command recognition device according to claim 2,

wherein the motion control instructing unit determines whether the command of the word sequence information should be executed based on the command estimation value which is calculated using the degree of speech confidence, the degree of image confidence, and the degree of motion confidence.

4. The command recognition device according to claim 1, further comprising attention determining unit that determines whether speech uttering person gives attention to a robot based on a captured image and that executes the command of the word sequence information when determining that the speech uttering person gives attention to the robot.

5. The command recognition device according to claim 1, wherein the utterance understanding unit determines command information based on word sequence information of which a command is previously determined as being executed.

6. A command recognition method in a command recognition device, comprising:

a first step of causing the command recognition device to determine or select word sequence information from speech information;

a second step of causing the command recognition device to calculate degree of speech confidence based on the speech information and the word sequence information;

a third step of causing the command recognition device to calculate a degree of phrase confidence based on image information and phrase information included in the word sequence information; and

a fourth step of causing the command recognition device to determine whether a command of the word sequence information should be executed based on a command estimation value calculated using the degree of speech confidence and the degree of phrase confidence,

wherein, in the third step, a trajectory-of-motion of an object is calculated from the image information, the trajectory of motion being calculated based on time-series data of position coordinates of a trajectory in a motion of an object, where the position coordinates are variable in time, and a degree of motion confidence representing the confidence that the trajectory-of-motion comprises a motion of the phrase information is calculated as the degree of phrase confidence.

7. A command recognition robot comprising:

an utterance understanding unit that determines or selects word sequence information from speech information;

speech confidence degree calculating unit that calculates degree of speech confidence based on the speech information and the word sequence information;

a phrase confidence degree calculating unit that calculates a degree of phrase confidence based on image information and phrase information included in the word sequence information;

an image analysis unit that calculates a trajectory-of-motion of an object from the image information, wherein the trajectory of motion is calculated based on time-series data of position coordinates of a trajectory in a motion of an object, where the position coordinates are variable in time;

a motion control instructing unit that determines whether a command of the word sequence information should be executed based on a command estimation value which is calculated using the degree of speech confidence and the degree of phrase confidence; and

a motion mechanism unit that executes the command based on the word sequence information,

wherein the phrase confidence degree calculating unit calculates a degree of motion confidence representing the confidence that the trajectory-of-motion comprises a motion of the phrase information as the degree of phrase confidence.