Robot and controlling method thereof

1. A robot comprising:

an output interface including at least one of a display and a speaker;

a camera;

at least one first motor configured to rotate the output interface;

at least one emotion recognizer trained to recognize emotion of a user from input data through deep learning; and

a processor configured to:

control the output interface to output content to the display or to the speaker,

control the camera to capture at least one image the user watching the content while the content is output,

detect an over-immersion state of the user based on an emotion result of the at least one emotion recognizer using the at least one image as the input data, and

when the over-immersion state is detected, release the over-immersion state of the user to control driving of the at least one first motor to rotate the output interface for releasing over-immersion of the user.

2. The robot of claim 1 , wherein the input data of the at least one emotion recognizer further comprises text data obtained from a speech of the user and audio data of the user including at least one of tone, volume, waveform, magnitude and pitch of the audio data.

3. The robot of claim 2 , further comprising a multi-modal emotion recognizer to output the emotion result through multi-modal deep learning based on an output of the at least one emotion recognizer,

wherein the emotion recognizer includes at least one recognizer implemented through deep learning.

4. The robot of claim 1 , wherein the processor periodically or continuously recognizes the emotion of the user from the captured at least one image through the at least one emotion recognizer, the at least one image includes a video or a plurality of still images captured at predetermined time points.

5. The robot of claim 4 , wherein the processor is further configured to:

detect that the user is in an over-immersion state of the content when the recognized emotion remains the same during a reference time.

6. The robot of claim 1 , wherein the processor is further configured to:

track a gaze of the user from the at least one image, and

detect that the user is in the over-immersion state based on a result of tracking.

7. The robot of claim 6 , wherein the processor detects that the user is in the over-immersion state, upon detecting that a position of the gaze is constant during a reference time.

8. The robot of claim 6 , wherein the output interface includes the display,

wherein the content includes a graphical user interface (GUI) output through the display, and

wherein the processor is further configured to:

acquire display position change information of the GUI,

track a gaze of the user to determine a gaze position change of the user, and

when a position change of the GUI based on the acquired display position change information corresponds to the gaze position change of the user, detect that the user is in the over-immersion state.

9. The robot of claim 1 , wherein the processor is further configured to:

measure a number of eye blinks of the user from the at least one image during a reference time, and

when the measured number of eye blinks is less than a reference number, determine that the user is in the over-immersion state.

10. The robot of claim 1 , further comprising a communication transceiver for connecting the robot to a server,

wherein the processor is further configured to:

transmit the at least one image to the server through the communication transceiver,

when the over-immersion state of the user is detected by the server, receive over-immersion detection information from the server, and

control the operation of releasing the over-immersion state of the user in response to the received over-immersion detection information.

11. The robot of claim 1 , wherein, wherein the output interface includes the display and the speaker, and

wherein the release of the over-immersion of the state of the user includes:

displaying a touch item on the display, and

outputting a message through the speaker for inducing touch by the user of the touch item.

12. The robot of claim 1 , further comprising at least one second motor configured to tilt the output interface of the robot,

wherein, when the over-immersion state is detected, the processor is further configured to control driving of the at least one second motor to tilt the output interface of the robot.

13. The robot of claim 1 , further comprising a communication transceiver for connection with a terminal,

wherein, when the over-immersion state is detected, the processor transmits, to the terminal, a notification indicating that the user is in an over-immersion state.

14. A method of controlling a robot including a camera and at least one of a display and a speaker, the method comprising:

outputting content through the at least one of the display and the speaker;

capturing at least one image of a user watching the content through the camera while the content is output;

detecting an over-immersion state of the user based on an emotion result of at least one emotion recognizer performing deep learning using the captured at least one image as input data; and

after the over-immersion state is detected, releasing over-immersion of the user through driving of at least one first motor to rotate the output interface or at least one second motor to tilt the output interface.

15. The method of claim 14 , wherein the detecting of the over-immersion state includes:

recognizing emotion of the user through the deep learning using at least one of the at least one captured image, audio data of the user and text data obtained from a speech of the user as the input data; and

recognizing the over-immersion state of the user based on the emotion result obtained through multi-modal deep learning based on an output of the deep learning.

16. The method of claim 15 , wherein the detecting of the over-immersion state based on the recognized emotion includes detecting that the user is in an over-immersion state, when the recognized emotion according to the emotion result remains the same during a reference time or more.

17. The method of claim 14 , wherein the detecting of the over-immersion state includes:

tracking a gaze of the user watching the content from the at least one image; and

detecting the over-immersion state based on a result of tracking.

18. The method of claim 17 , wherein the detecting of the over-immersion state based on the result of tracking includes detecting that the user is in an over-immersion state, upon detecting that a position of the gaze of the user is constant during a reference time.

19. The method of claim 17 , wherein the output interface includes the display,

wherein the content includes a graphical user interface (GUI) output through the display, and

wherein the detecting of the over-immersion state based on the result of tracking includes:

acquiring display position change information of the GUI;

comparing a position change of the GUI based on the acquired display position change information with a gaze position change of the user detected according to the tracking of the gaze; and

when the position change of the GUI corresponds to the gaze position change during a reference time as a result of comparison, detecting that the user is in the over-immersion state.

20. The method of claim 14 , wherein the controlling of operation of releasing the over-immersion state further includes at least one of:

displaying a touch item through the display and outputting a message through the speaker for inducing touch of the touch item by the user;

inducing the user to perform an operation which is not related to the content or

transmitting, to a terminal, a notification indicating that the user is in an over-immersion state.