Artificial intelligence robot and method of operating the same

Patent metadata

Field	Value
Publication number	US-12096895-B2
Application number	US-201916963822-A
Country	US
Kind code	B2
Filing date	Jun 14, 2019
Priority date	Jun 14, 2019
Publication date	Sep 24, 2024
Grant date	Sep 24, 2024

Abstract

An artificial intelligence robot includes a camera configured to acquire image data, a memory configured to store an object recognition model used to recognize an object from the image data, and a processor configured to acquire a speech command, determine whether an intention of the acquired speech command is object search, recognize an object from the image data based on the object recognition model during traveling when the intention of object search, and output a notification indicating that the object has been recognized when the recognized object is an intended object.

First claim

The invention claimed is:

1. An artificial intelligence robot comprising:
a camera configured to acquire image data;
a memory configured to store an object recognition model; and
a processor configured to:
acquire a speech command;
determine whether an intention of the acquired speech command is object search;
recognize an object from the image data based on the object recognition model during traveling according to a first cleaning route when the intention represents the object search;
output a notification indicating that the object has been recognized when the recognized object is an intended object;
identify a position of the recognized object on a cleaning map including a plurality of cleaning areas when the object is recognized from the image data; and
change the first cleaning route to a second cleaning route such that a cleaning area including the position of the object on the first cleaning route is removed,
wherein the processor is further configured to:
acquire identification information of the intended object; and
determine that the intended object has been searched when the acquired identification information of the object matches identification information of the recognized object.

2. The artificial intelligence robot according to claim 1 , wherein the identification information of the object includes a name or type of the object.

3. The artificial intelligence robot according to claim 2 , further comprising a speaker configured to output audio,
wherein the processor outputs the notification through the speaker.

4. The artificial intelligence robot according to claim 1 , further comprising a short-range wireless communication module configured to wirelessly communicate with a mobile terminal,
wherein the processor transmits the notification to the mobile terminal through the short-range wireless communication module, and
wherein the notification includes position information indicating the position of the recognized object on the cleaning map.

5. The artificial intelligence robot according to claim 1 , wherein the object recognition model is an artificial neural network model learned by a deep learning algorithm or a machine learning algorithm.

6. The artificial intelligence robot according to claim 5 , wherein the object recognition model is learned through supervised learning and is a model for determining identification information of an object using image data for learning and object identification data labeled with the image data for learning.

7. A method of operating an artificial intelligence robot, the method comprising:
acquiring a speech command;
determining that an intention of the acquired speech command is object search;
acquiring image data based on determining that the intention is object search;
recognizing an object from the image data based on an object recognition model during raveling according to a first cleaning route, based on the intention representing the object search;
outputting a notification indicating that the object has been recognized, based on the recognized object being an intended object;
identifying a position of the recognized object on a cleaning map including a plurality of cleaning areas, based on the object being recognized from the image data;
changing the first cleaning route to a second cleaning route such that a cleaning area including the position of the object on the first cleaning route is removed;
acquiring identification information of the intended object; and
determining that the intended object has been searched, based on the acquired identification information of the object matching identification information of the recognized object.

8. The method according to claim 7 ,
wherein the identification information of the object includes a name or type of the object.

9. The method according to claim 7 , further comprising transmitting the notification to a mobile terminal, and
wherein the notification includes position information indicating the position of the recognized object on the cleaning map.

10. The method according to claim 7 , wherein the object recognition model is an artificial neural network model learned by a deep learning algorithm or a machine learning algorithm, is learned through supervised learning, and is a model for determining identification information of an object using image data for learning and object identification data labeled with the image data for learning.

Assignees

• Lg Electronics Inc

Inventors

• Chae Jonghoon
• Kim Beomoh
• Kim Taehyun
• Maeng Jichan
• Shin Wonho

Classifications

• G06N3/0464

  Convolutional networks [CNN, ConvNet] · CPC title

• G06N3/09

  Supervised learning · CPC title

• B25J19/023

  including video camera means · CPC title

• B25J13/006

  by means of a wireless system for controlling one or several manipulators · CPC title

• B25J13/003

  by means of an audio-responsive input (audible safety signals B25J19/061) · CPC title