# US 2010/0138040 A1 · Apparatus for detecting user and method for detecting user by the same

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2010/0138040 A1
- Document ID: US20100138040A1
- Status: Published application
- Kind: A1
- Application number: 12/373,141
- Filed: 2008-01-14
- Published: 2010-06-03
- Earliest claimed: 2007-01-18
- Assignees: Korea Institute of Science and Technology
- Inventors: Soon-Il Kwon
- Web page: https://patentgrep.com/patent/US20100138040A1

## CPC

- B25J 13/003: {by means of an audio-responsive input (audible safety signals B25J19/061)}
- G10L 17/06: Decision making techniques; Pattern matching strategies
- B25J: MANIPULATORS; CHAMBERS PROVIDED WITH MANIPULATION DEVICES ({manipulators specially adapted for use in surgery A61B34/70; manipulators used in cleaning hollow articles B08B9/04}; manipulators associated with rolling mills B21B39/20; manipulators associated with forging machines B21J13/10; {manipulators associated with picking-up and placing mechanisms B23P19/007}; means for holding wheels or parts thereof B60B30/00; {vehicles with ground-engaging propulsion means, e.g. walking members B62D57/02, B62D57/032; devices for picking-up and depositing articles or materials between conveyors B65G47/90, B65G47/91; manipulators with gripping or holding means for transferring packages B65H67/065}; cranes B66C; {manipulators used in the protection or supervision of pipe-line installations F17D5/00; walking equipment adapted for nuclear steam-generators F22B37/006}; manipulators specially adapted for, or associated with, nuclear reactors G21C; {apparatus used for handling wafers during manufacture or treatment of semiconductor H10P72/50})
- G10L: SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING

## Abstract

An apparatus for identifying a user in a service robot for domestic use is provided, the apparatus comprising a feature vector classifying unit for classifying feature vectors converted from a user voice signal for statistics into a feature vector not causing a user identification error and a feature vector causing the user identification error to thereby set user modes corresponding to the respective feature vectors for each user, and an identifying unit for identifying the user by utilizing the classification result of the feature vector classifying unit as data for the user identification. Accordingly, time restriction caused when the service robot for domestic use identifies who a user is can be overcome, and, the degradation of the user identification due to the lack of absolutely needed data amount when the user is identified only by the user's short voice can be solved.

## Claims

### Claim 1

An apparatus for identifying a user comprising:
a feature vector classifying unit configured to compare a real voice signal of a user with preset user models when the real voice of the user is inputted, and to classify feature vectors of the real voice signal into a first-type feature vector not causing a user identification error and a second-type feature vector causing the user identification error; and
a identifying unit configured to identify who a user is, by using the classification result of the feature vector classifying unit as data for user identification, wherein the identifying unit identifies the user by only using the first-type feature vector not causing the user identification error.

### Claim 2 (depends on 1)

The apparatus of claim 1, wherein the apparatus for identifying the user is applied to a service robot for domestic use.

### Claim 3 (depends on 1)

The apparatus of claim 1, wherein the feature vector classifying unit compares at least one of the first-type and second-type feature vectors of the real voice signal with the user models, thereby to classify the at least one feature vector into a feature vector with a non-overlapped characteristic in statistical models and a feature vector with an overlapped characteristic.

### Claim 4 (depends on 3)

The apparatus of claim 3, wherein the identifying unit compares a feature vector, which has the non-overlapped characteristic in the statistical models and does not cause the user identification error, with the user models.

### Claim 5

A service robot for domestic use comprising:
an input unit configured to receive external sound;
a feature vector extracting unit configured to receive a signal from the input unit and convert the received signal into feature vectors;
a feature vector classifying unit configured to classify the feature vectors for the external sound according to user models set based upon the feature vectors converted from a user voice signal for statistics;
an identifying unit configured to identify who a user is, only by utilizing a feature vector not causing a user identification error among feature vectors classified by the feature vector classifying unit; and
a controlling unit configured to control a command executing unit according to the determination result of the identifying unit.

### Claim 6 (depends on 5)

The service robot of claim 5, wherein the external sound is the user's voice signal.

### Claim 7 (depends on 5)

The service robot of claim 5, wherein the input unit comprises a microphone.

### Claim 8 (depends on 5)

The service robot of claim 5, wherein the feature vector classifying unit classifies feature vectors corresponding to the external sounds into a feature vector with a non-overlapped characteristic in statistical models and a feature vector with an overlapped characteristic in the statistical models.

### Claim 9 (depends on 8)

The service robot of claim 8, wherein the identifying unit uses only the feature vector of the external sound having the non-overlapped characteristic in the statistical models, to compare the feature vector of the external sound having the non-overlapped characteristic in the statistical models with the user models according to a maximum likelihood method.

### Claim 10

A method for identifying who a user is, by a service robot for domestic, use comprising:
receiving voice;
extracting feature vectors from a signal of the voice;
comparing the feature vectors with preset user models to classify the feature vectors into a first-type feature vector not causing a user identification error and a second-type feature vector causing the user identification error; and
comparing the first-type feature vector with the user models to identify who the user is.

### Claim 11 (depends on 10)

The method of claim 10, wherein a maximum likelihood method is applied to compare the feature vector with the user models.

### Claim 12 (depends on 10)

The method of claim 10, wherein a maximum likelihood method is applied to compare the first-type feature vector with the user models.
