Ensemble Learning with Weak Classifiers for Fast and Reliable Unknown Terrain Classification Using Mobile Robots

Research output: Contribution to journalArticle

7 Citations (Scopus)

Abstract

We propose a lightweight and fast learning algorithm for classifying the features of an unknown terrain that a robot is navigating in. Most of the existing research on unknown terrain classification by mobile robots relies on a single powerful classifier to correctly identify the terrain using sensor data from a single sensor like laser or camera. In contrast, our proposed approach uses multiple modalities of sensed data and multiple, weak but less-complex classifiers for classifying the terrain types. The classifiers are combined using an ensemble learning algorithm to improve the algorithm's training rate as compared to an individual classifier. Our algorithm was tested with data collected by navigating a four-wheeled, autonomous robot, called Explorer, over different terrains including brick, grass, rock, sand, and concrete. Our results show that our proposed approach performs better with up to 63% better prediction accuracy for some terrains as compared to a support vector machine (SVM)-based learning technique that uses sensor data from a single sensor. Despite using multiple classifiers, our algorithm takes only a fraction (1/65) of the time on average, as compared to the SVM technique.

Original languageEnglish (US)
Article number7464340
Pages (from-to)2933-2944
Number of pages12
JournalIEEE Transactions on Systems, Man, and Cybernetics: Systems
Volume47
Issue number11
DOIs
StatePublished - Nov 2017

Fingerprint

Mobile robots
Classifiers
Sensors
Learning algorithms
Support vector machines
Robots
Brick
Sand
Cameras
Rocks
Concretes
Lasers

Keywords

  • Ensemble learning
  • mobile robot
  • terrain classification

ASJC Scopus subject areas

  • Software
  • Control and Systems Engineering
  • Human-Computer Interaction
  • Computer Science Applications
  • Electrical and Electronic Engineering

Cite this

@article{4798fa6109b64d0ca228b6e337869251,
title = "Ensemble Learning with Weak Classifiers for Fast and Reliable Unknown Terrain Classification Using Mobile Robots",
abstract = "We propose a lightweight and fast learning algorithm for classifying the features of an unknown terrain that a robot is navigating in. Most of the existing research on unknown terrain classification by mobile robots relies on a single powerful classifier to correctly identify the terrain using sensor data from a single sensor like laser or camera. In contrast, our proposed approach uses multiple modalities of sensed data and multiple, weak but less-complex classifiers for classifying the terrain types. The classifiers are combined using an ensemble learning algorithm to improve the algorithm's training rate as compared to an individual classifier. Our algorithm was tested with data collected by navigating a four-wheeled, autonomous robot, called Explorer, over different terrains including brick, grass, rock, sand, and concrete. Our results show that our proposed approach performs better with up to 63{\%} better prediction accuracy for some terrains as compared to a support vector machine (SVM)-based learning technique that uses sensor data from a single sensor. Despite using multiple classifiers, our algorithm takes only a fraction (1/65) of the time on average, as compared to the SVM technique.",
keywords = "Ensemble learning, mobile robot, terrain classification",
author = "Ayan Dutta and Prithviraj Dasgupta",
year = "2017",
month = "11",
doi = "10.1109/TSMC.2016.2531700",
language = "English (US)",
volume = "47",
pages = "2933--2944",
journal = "IEEE Transactions on Systems, Man, and Cybernetics: Systems",
issn = "2168-2216",
publisher = "IEEE Advancing Technology for Humanity",
number = "11",

}

TY - JOUR

T1 - Ensemble Learning with Weak Classifiers for Fast and Reliable Unknown Terrain Classification Using Mobile Robots

AU - Dutta, Ayan

AU - Dasgupta, Prithviraj

PY - 2017/11

Y1 - 2017/11

N2 - We propose a lightweight and fast learning algorithm for classifying the features of an unknown terrain that a robot is navigating in. Most of the existing research on unknown terrain classification by mobile robots relies on a single powerful classifier to correctly identify the terrain using sensor data from a single sensor like laser or camera. In contrast, our proposed approach uses multiple modalities of sensed data and multiple, weak but less-complex classifiers for classifying the terrain types. The classifiers are combined using an ensemble learning algorithm to improve the algorithm's training rate as compared to an individual classifier. Our algorithm was tested with data collected by navigating a four-wheeled, autonomous robot, called Explorer, over different terrains including brick, grass, rock, sand, and concrete. Our results show that our proposed approach performs better with up to 63% better prediction accuracy for some terrains as compared to a support vector machine (SVM)-based learning technique that uses sensor data from a single sensor. Despite using multiple classifiers, our algorithm takes only a fraction (1/65) of the time on average, as compared to the SVM technique.

AB - We propose a lightweight and fast learning algorithm for classifying the features of an unknown terrain that a robot is navigating in. Most of the existing research on unknown terrain classification by mobile robots relies on a single powerful classifier to correctly identify the terrain using sensor data from a single sensor like laser or camera. In contrast, our proposed approach uses multiple modalities of sensed data and multiple, weak but less-complex classifiers for classifying the terrain types. The classifiers are combined using an ensemble learning algorithm to improve the algorithm's training rate as compared to an individual classifier. Our algorithm was tested with data collected by navigating a four-wheeled, autonomous robot, called Explorer, over different terrains including brick, grass, rock, sand, and concrete. Our results show that our proposed approach performs better with up to 63% better prediction accuracy for some terrains as compared to a support vector machine (SVM)-based learning technique that uses sensor data from a single sensor. Despite using multiple classifiers, our algorithm takes only a fraction (1/65) of the time on average, as compared to the SVM technique.

KW - Ensemble learning

KW - mobile robot

KW - terrain classification

UR - http://www.scopus.com/inward/record.url?scp=85035784677&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85035784677&partnerID=8YFLogxK

U2 - 10.1109/TSMC.2016.2531700

DO - 10.1109/TSMC.2016.2531700

M3 - Article

AN - SCOPUS:85035784677

VL - 47

SP - 2933

EP - 2944

JO - IEEE Transactions on Systems, Man, and Cybernetics: Systems

JF - IEEE Transactions on Systems, Man, and Cybernetics: Systems

SN - 2168-2216

IS - 11

M1 - 7464340

ER -