Learning to fire at targets by an iCub humanoid robot

Vishnu K. Nath, Stephen E Levinson

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

In this paper, we present an algorithm that integrates computer vision with machine learning to enable a humanoid robot to accurately fire at objects classified as targets. The robot needs to be calibrated to hold the gun and instructed how to pull the trigger. Two algorithms are proposed and are executed depending on the dynamics of the target. If the target is stationery, a least mean square (LMS) approach is used to computethe error and adjust the gun muzzle accordingly. If the target is found to be dynamic, a modified Q-learning is used to best predict the object position and velocity and to adjust relevant parameters, as necessary. The image processing utilizes the OpenCV library to detect the target and point of impact of the bullets. The approach is evaluated on a 53-DOF humanoid robot iCub. This work is an example of fine motor control which is the basis for much of natural language processing by spatial reasoning. It is one aspect of a long term research effort on automatic language acquisition [3].

Original languageEnglish (US)
Title of host publicationDesigning Intelligent Robots
Subtitle of host publicationReintegrating AI II - Papers from the AAAI Spring Symposium, Technical Report
Pages33-36
Number of pages4
StatePublished - Sep 5 2013
Event2013 AAAI Spring Symposium - Palo Alto, CA, United States
Duration: Mar 25 2013Mar 27 2013

Publication series

NameAAAI Spring Symposium - Technical Report
VolumeSS-13-04

Other

Other2013 AAAI Spring Symposium
CountryUnited States
CityPalo Alto, CA
Period3/25/133/27/13

Fingerprint

Fires
Robots
Computer vision
Learning systems
Image processing
Processing

ASJC Scopus subject areas

  • Artificial Intelligence

Cite this

Nath, V. K., & Levinson, S. E. (2013). Learning to fire at targets by an iCub humanoid robot. In Designing Intelligent Robots: Reintegrating AI II - Papers from the AAAI Spring Symposium, Technical Report (pp. 33-36). (AAAI Spring Symposium - Technical Report; Vol. SS-13-04).

Learning to fire at targets by an iCub humanoid robot. / Nath, Vishnu K.; Levinson, Stephen E.

Designing Intelligent Robots: Reintegrating AI II - Papers from the AAAI Spring Symposium, Technical Report. 2013. p. 33-36 (AAAI Spring Symposium - Technical Report; Vol. SS-13-04).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Nath, VK & Levinson, SE 2013, Learning to fire at targets by an iCub humanoid robot. in Designing Intelligent Robots: Reintegrating AI II - Papers from the AAAI Spring Symposium, Technical Report. AAAI Spring Symposium - Technical Report, vol. SS-13-04, pp. 33-36, 2013 AAAI Spring Symposium, Palo Alto, CA, United States, 3/25/13.
Nath VK, Levinson SE. Learning to fire at targets by an iCub humanoid robot. In Designing Intelligent Robots: Reintegrating AI II - Papers from the AAAI Spring Symposium, Technical Report. 2013. p. 33-36. (AAAI Spring Symposium - Technical Report).
Nath, Vishnu K. ; Levinson, Stephen E. / Learning to fire at targets by an iCub humanoid robot. Designing Intelligent Robots: Reintegrating AI II - Papers from the AAAI Spring Symposium, Technical Report. 2013. pp. 33-36 (AAAI Spring Symposium - Technical Report).
@inproceedings{dbe84114af014c77aa502e7adfc65c50,
title = "Learning to fire at targets by an iCub humanoid robot",
abstract = "In this paper, we present an algorithm that integrates computer vision with machine learning to enable a humanoid robot to accurately fire at objects classified as targets. The robot needs to be calibrated to hold the gun and instructed how to pull the trigger. Two algorithms are proposed and are executed depending on the dynamics of the target. If the target is stationery, a least mean square (LMS) approach is used to computethe error and adjust the gun muzzle accordingly. If the target is found to be dynamic, a modified Q-learning is used to best predict the object position and velocity and to adjust relevant parameters, as necessary. The image processing utilizes the OpenCV library to detect the target and point of impact of the bullets. The approach is evaluated on a 53-DOF humanoid robot iCub. This work is an example of fine motor control which is the basis for much of natural language processing by spatial reasoning. It is one aspect of a long term research effort on automatic language acquisition [3].",
author = "Nath, {Vishnu K.} and Levinson, {Stephen E}",
year = "2013",
month = "9",
day = "5",
language = "English (US)",
isbn = "9781577356011",
series = "AAAI Spring Symposium - Technical Report",
pages = "33--36",
booktitle = "Designing Intelligent Robots",

}

TY - GEN

T1 - Learning to fire at targets by an iCub humanoid robot

AU - Nath, Vishnu K.

AU - Levinson, Stephen E

PY - 2013/9/5

Y1 - 2013/9/5

N2 - In this paper, we present an algorithm that integrates computer vision with machine learning to enable a humanoid robot to accurately fire at objects classified as targets. The robot needs to be calibrated to hold the gun and instructed how to pull the trigger. Two algorithms are proposed and are executed depending on the dynamics of the target. If the target is stationery, a least mean square (LMS) approach is used to computethe error and adjust the gun muzzle accordingly. If the target is found to be dynamic, a modified Q-learning is used to best predict the object position and velocity and to adjust relevant parameters, as necessary. The image processing utilizes the OpenCV library to detect the target and point of impact of the bullets. The approach is evaluated on a 53-DOF humanoid robot iCub. This work is an example of fine motor control which is the basis for much of natural language processing by spatial reasoning. It is one aspect of a long term research effort on automatic language acquisition [3].

AB - In this paper, we present an algorithm that integrates computer vision with machine learning to enable a humanoid robot to accurately fire at objects classified as targets. The robot needs to be calibrated to hold the gun and instructed how to pull the trigger. Two algorithms are proposed and are executed depending on the dynamics of the target. If the target is stationery, a least mean square (LMS) approach is used to computethe error and adjust the gun muzzle accordingly. If the target is found to be dynamic, a modified Q-learning is used to best predict the object position and velocity and to adjust relevant parameters, as necessary. The image processing utilizes the OpenCV library to detect the target and point of impact of the bullets. The approach is evaluated on a 53-DOF humanoid robot iCub. This work is an example of fine motor control which is the basis for much of natural language processing by spatial reasoning. It is one aspect of a long term research effort on automatic language acquisition [3].

UR - http://www.scopus.com/inward/record.url?scp=84883291616&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84883291616&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:84883291616

SN - 9781577356011

T3 - AAAI Spring Symposium - Technical Report

SP - 33

EP - 36

BT - Designing Intelligent Robots

ER -