Online learning for offroad robots

Using spatial label propagation to learn long-range traversability

Raia Hadsell, Pierre Sermanet, Ayse Naz Erkan, Jan Ben, Jefferson Han, Beat Flepp, Urs Muller, Yann LeCun

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

Abstract

We present a solution to the problem of long-range obstacle/path recognition in autonomous robots. The system uses sparse traversability information from a stereo module to train a classifier online. The trained classifier can then predict the traversability of the entire scene. A distance-normalized image pyramid makes it possible to efficiently train on each frame seen by the robot, using large windows that contain contextual information as well as shape, color, and texture. Traversability labels are initially obtained for each target using a stereo module, then propagated to other views of the same target using temporal and spatial concurrences, thus training the classifier to be viewinvariant. A ring buffer simulates short-term memory and ensures that the discriminative learning is balanced and consistent. This long-range obstacle detection system sees obstacles and paths at 30-40 meters, far beyond the maximum stereo range of 12 meters, and adapts very quickly to new environments. Experiments were run on the LAGR robot platform.

Original languageEnglish (US)
Title of host publicationRobotics: Science and Systems III
PublisherMIT Press Journals
Pages17-23
Number of pages7
Volume3
ISBN (Print)9780262524841
StatePublished - 2008
Event3rd International Conference on Robotics Science and Systems, RSS 2007 - Atlanta, United States
Duration: Jun 27 2007Jun 30 2007

Other

Other3rd International Conference on Robotics Science and Systems, RSS 2007
CountryUnited States
CityAtlanta
Period6/27/076/30/07

Fingerprint

Labels
Classifiers
Robots
Textures
Color
Data storage equipment
Experiments

ASJC Scopus subject areas

  • Artificial Intelligence
  • Control and Systems Engineering
  • Electrical and Electronic Engineering

Cite this

Hadsell, R., Sermanet, P., Erkan, A. N., Ben, J., Han, J., Flepp, B., ... LeCun, Y. (2008). Online learning for offroad robots: Using spatial label propagation to learn long-range traversability. In Robotics: Science and Systems III (Vol. 3, pp. 17-23). MIT Press Journals.

Online learning for offroad robots : Using spatial label propagation to learn long-range traversability. / Hadsell, Raia; Sermanet, Pierre; Erkan, Ayse Naz; Ben, Jan; Han, Jefferson; Flepp, Beat; Muller, Urs; LeCun, Yann.

Robotics: Science and Systems III. Vol. 3 MIT Press Journals, 2008. p. 17-23.

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

Hadsell, R, Sermanet, P, Erkan, AN, Ben, J, Han, J, Flepp, B, Muller, U & LeCun, Y 2008, Online learning for offroad robots: Using spatial label propagation to learn long-range traversability. in Robotics: Science and Systems III. vol. 3, MIT Press Journals, pp. 17-23, 3rd International Conference on Robotics Science and Systems, RSS 2007, Atlanta, United States, 6/27/07.
Hadsell R, Sermanet P, Erkan AN, Ben J, Han J, Flepp B et al. Online learning for offroad robots: Using spatial label propagation to learn long-range traversability. In Robotics: Science and Systems III. Vol. 3. MIT Press Journals. 2008. p. 17-23
Hadsell, Raia ; Sermanet, Pierre ; Erkan, Ayse Naz ; Ben, Jan ; Han, Jefferson ; Flepp, Beat ; Muller, Urs ; LeCun, Yann. / Online learning for offroad robots : Using spatial label propagation to learn long-range traversability. Robotics: Science and Systems III. Vol. 3 MIT Press Journals, 2008. pp. 17-23
@inproceedings{9dda43834055424db7633973382ecfe4,
title = "Online learning for offroad robots: Using spatial label propagation to learn long-range traversability",
abstract = "We present a solution to the problem of long-range obstacle/path recognition in autonomous robots. The system uses sparse traversability information from a stereo module to train a classifier online. The trained classifier can then predict the traversability of the entire scene. A distance-normalized image pyramid makes it possible to efficiently train on each frame seen by the robot, using large windows that contain contextual information as well as shape, color, and texture. Traversability labels are initially obtained for each target using a stereo module, then propagated to other views of the same target using temporal and spatial concurrences, thus training the classifier to be viewinvariant. A ring buffer simulates short-term memory and ensures that the discriminative learning is balanced and consistent. This long-range obstacle detection system sees obstacles and paths at 30-40 meters, far beyond the maximum stereo range of 12 meters, and adapts very quickly to new environments. Experiments were run on the LAGR robot platform.",
author = "Raia Hadsell and Pierre Sermanet and Erkan, {Ayse Naz} and Jan Ben and Jefferson Han and Beat Flepp and Urs Muller and Yann LeCun",
year = "2008",
language = "English (US)",
isbn = "9780262524841",
volume = "3",
pages = "17--23",
booktitle = "Robotics: Science and Systems III",
publisher = "MIT Press Journals",

}

TY - GEN

T1 - Online learning for offroad robots

T2 - Using spatial label propagation to learn long-range traversability

AU - Hadsell, Raia

AU - Sermanet, Pierre

AU - Erkan, Ayse Naz

AU - Ben, Jan

AU - Han, Jefferson

AU - Flepp, Beat

AU - Muller, Urs

AU - LeCun, Yann

PY - 2008

Y1 - 2008

N2 - We present a solution to the problem of long-range obstacle/path recognition in autonomous robots. The system uses sparse traversability information from a stereo module to train a classifier online. The trained classifier can then predict the traversability of the entire scene. A distance-normalized image pyramid makes it possible to efficiently train on each frame seen by the robot, using large windows that contain contextual information as well as shape, color, and texture. Traversability labels are initially obtained for each target using a stereo module, then propagated to other views of the same target using temporal and spatial concurrences, thus training the classifier to be viewinvariant. A ring buffer simulates short-term memory and ensures that the discriminative learning is balanced and consistent. This long-range obstacle detection system sees obstacles and paths at 30-40 meters, far beyond the maximum stereo range of 12 meters, and adapts very quickly to new environments. Experiments were run on the LAGR robot platform.

AB - We present a solution to the problem of long-range obstacle/path recognition in autonomous robots. The system uses sparse traversability information from a stereo module to train a classifier online. The trained classifier can then predict the traversability of the entire scene. A distance-normalized image pyramid makes it possible to efficiently train on each frame seen by the robot, using large windows that contain contextual information as well as shape, color, and texture. Traversability labels are initially obtained for each target using a stereo module, then propagated to other views of the same target using temporal and spatial concurrences, thus training the classifier to be viewinvariant. A ring buffer simulates short-term memory and ensures that the discriminative learning is balanced and consistent. This long-range obstacle detection system sees obstacles and paths at 30-40 meters, far beyond the maximum stereo range of 12 meters, and adapts very quickly to new environments. Experiments were run on the LAGR robot platform.

UR - http://www.scopus.com/inward/record.url?scp=84959292080&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84959292080&partnerID=8YFLogxK

M3 - Conference contribution

SN - 9780262524841

VL - 3

SP - 17

EP - 23

BT - Robotics: Science and Systems III

PB - MIT Press Journals

ER -