Behavior recognition for humanoid robots using long short-term memory

Research output: Contribution to journalArticle

6 Citations (Scopus)

Abstract

Learning from demonstration plays an important role in enabling robot to acquire new behaviors from human teachers. Within learning from demonstration, robots learn new tasks by recognizing a set of preprogrammed behaviors or skills as building blocks for new, potentially more complex tasks. One important aspect in this approach is the recognition of the set of behaviors that comprises the entire task. The ability to recognize a complex task as a sequence of simple behaviors enables the robot to generalize better on more complex tasks. In this article, we propose that primitive behaviors can be taught to a robot via learning from demonstration. In our experiment, we teach the robot new behaviors by demonstrating the behaviors to the robot several times. Following that, a long short-term memory recurrent neural network is trained to recognize the behaviors. In this study, we managed to teach at least six behaviors on a NAO humanoid robot and trained a long short-term memory recurrent neural network to recognize the behaviors using the supervised learning scheme. Our result shows that long short-term memory can recognize all the taught behaviors effectively, and it is able to generalize to recognize similar types of behaviors that have not been demonstrated on the robot before. We also show that the long short-term memory is advantageous compared to other neural network frameworks in recognizing the behaviors in the presence of noise in the behaviors.

Original languageEnglish
Pages (from-to)1-14
Number of pages14
JournalInternational Journal of Advanced Robotic Systems
Volume13
Issue number6
DOIs
Publication statusPublished - 26 Oct 2016

Fingerprint

Robots
Demonstrations
Recurrent neural networks
Long short-term memory
Supervised learning
Neural networks
Experiments

All Science Journal Classification (ASJC) codes

  • Software
  • Computer Science Applications
  • Artificial Intelligence

Cite this

@article{39c51b99e4134ac180a434a3913f9113,
title = "Behavior recognition for humanoid robots using long short-term memory",
abstract = "Learning from demonstration plays an important role in enabling robot to acquire new behaviors from human teachers. Within learning from demonstration, robots learn new tasks by recognizing a set of preprogrammed behaviors or skills as building blocks for new, potentially more complex tasks. One important aspect in this approach is the recognition of the set of behaviors that comprises the entire task. The ability to recognize a complex task as a sequence of simple behaviors enables the robot to generalize better on more complex tasks. In this article, we propose that primitive behaviors can be taught to a robot via learning from demonstration. In our experiment, we teach the robot new behaviors by demonstrating the behaviors to the robot several times. Following that, a long short-term memory recurrent neural network is trained to recognize the behaviors. In this study, we managed to teach at least six behaviors on a NAO humanoid robot and trained a long short-term memory recurrent neural network to recognize the behaviors using the supervised learning scheme. Our result shows that long short-term memory can recognize all the taught behaviors effectively, and it is able to generalize to recognize similar types of behaviors that have not been demonstrated on the robot before. We also show that the long short-term memory is advantageous compared to other neural network frameworks in recognizing the behaviors in the presence of noise in the behaviors.",
author = "{Dickson Neoh}, {Tze How} and Loo, {Chu Kiong} and {Mohamed Sahari}, {Khairul Salleh}",
year = "2016",
month = "10",
day = "26",
doi = "10.1177/1729881416663369",
language = "English",
volume = "13",
pages = "1--14",
journal = "International Journal of Advanced Robotic Systems",
issn = "1729-8806",
publisher = "Vienna University of Technology",
number = "6",

}

Behavior recognition for humanoid robots using long short-term memory. / Dickson Neoh, Tze How; Loo, Chu Kiong; Mohamed Sahari, Khairul Salleh.

In: International Journal of Advanced Robotic Systems, Vol. 13, No. 6, 26.10.2016, p. 1-14.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Behavior recognition for humanoid robots using long short-term memory

AU - Dickson Neoh, Tze How

AU - Loo, Chu Kiong

AU - Mohamed Sahari, Khairul Salleh

PY - 2016/10/26

Y1 - 2016/10/26

N2 - Learning from demonstration plays an important role in enabling robot to acquire new behaviors from human teachers. Within learning from demonstration, robots learn new tasks by recognizing a set of preprogrammed behaviors or skills as building blocks for new, potentially more complex tasks. One important aspect in this approach is the recognition of the set of behaviors that comprises the entire task. The ability to recognize a complex task as a sequence of simple behaviors enables the robot to generalize better on more complex tasks. In this article, we propose that primitive behaviors can be taught to a robot via learning from demonstration. In our experiment, we teach the robot new behaviors by demonstrating the behaviors to the robot several times. Following that, a long short-term memory recurrent neural network is trained to recognize the behaviors. In this study, we managed to teach at least six behaviors on a NAO humanoid robot and trained a long short-term memory recurrent neural network to recognize the behaviors using the supervised learning scheme. Our result shows that long short-term memory can recognize all the taught behaviors effectively, and it is able to generalize to recognize similar types of behaviors that have not been demonstrated on the robot before. We also show that the long short-term memory is advantageous compared to other neural network frameworks in recognizing the behaviors in the presence of noise in the behaviors.

AB - Learning from demonstration plays an important role in enabling robot to acquire new behaviors from human teachers. Within learning from demonstration, robots learn new tasks by recognizing a set of preprogrammed behaviors or skills as building blocks for new, potentially more complex tasks. One important aspect in this approach is the recognition of the set of behaviors that comprises the entire task. The ability to recognize a complex task as a sequence of simple behaviors enables the robot to generalize better on more complex tasks. In this article, we propose that primitive behaviors can be taught to a robot via learning from demonstration. In our experiment, we teach the robot new behaviors by demonstrating the behaviors to the robot several times. Following that, a long short-term memory recurrent neural network is trained to recognize the behaviors. In this study, we managed to teach at least six behaviors on a NAO humanoid robot and trained a long short-term memory recurrent neural network to recognize the behaviors using the supervised learning scheme. Our result shows that long short-term memory can recognize all the taught behaviors effectively, and it is able to generalize to recognize similar types of behaviors that have not been demonstrated on the robot before. We also show that the long short-term memory is advantageous compared to other neural network frameworks in recognizing the behaviors in the presence of noise in the behaviors.

UR - http://www.scopus.com/inward/record.url?scp=85007281990&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85007281990&partnerID=8YFLogxK

U2 - 10.1177/1729881416663369

DO - 10.1177/1729881416663369

M3 - Article

AN - SCOPUS:85007281990

VL - 13

SP - 1

EP - 14

JO - International Journal of Advanced Robotic Systems

JF - International Journal of Advanced Robotic Systems

SN - 1729-8806

IS - 6

ER -