Toward bootstrap learning for place recognition
Benjamin Kuipers and Patrick Beeson. Toward bootstrap learning for place recognition. In Symposium on Anchoring Symbols to Sensory Data in Single and Multiple Robot Systems, AAAI Fall Symposium Series, North Falmouth, MA, November 2001. AAAI Technical Report FS-01-01.
Abstract
We present a method whereby a robot with no prior knowledge of its sensors, effectors or environment can learn to recognize places with high accuracy, in spite of perceptual aliasing (different places appear the same) and image variability (the same place appears differently). Previous work showed how such a robot could learn from its experience a useful set of sensory features, motion primitives, and local control laws to move from one distinctive state to another. Such progressive learning of a hierarchical representation is called bootstrap learning. The first step in learning place recognition eliminates image variability in two steps: (a) focusing on recognition of distinctive states defined by the robot's control laws, and (b) unsupervised learning of clusters of similar sensory images. The clusters define views associated with distinctive states, often increasing perceptual aliasing. The second step eliminates perceptual aliasing by building a cognitive map and using history information gathered during exploration to disambiguate distinctive states. The third step uses the labeled images for supervised learning of direct associations from sensory images to distinctive states. We evaluate the method using a physical mobile robot in two environments, showing large amounts of perceptual aliasing and high resulting recognition rates.
Additional Information
Also see: Benjamin Kuipers and Patrick Beeson. Bootstrap learning for place recognition. AAAI Conference on Artificial Intelligence, 2002.
BibTeX
@InProceedings{Kuipers-fss-01,
author = {Benjamin Kuipers and Patrick Beeson},
title = {Toward bootstrap learning for place recognition},
booktitle = {Symposium on Anchoring Symbols to Sensory Data in
Single and Multiple Robot Systems},
year = 2001,
series = {AAAI Fall Symposium Series},
address = {North Falmouth, MA},
month = {November},
note = {AAAI Technical Report FS-01-01.},
abstract = {We present a method whereby a robot with no prior
knowledge of its sensors, effectors or environment
can learn to recognize places with high accuracy, in
spite of perceptual aliasing (different places
appear the same) and image variability (the same
place appears differently). Previous work showed how
such a robot could learn from its experience a
useful set of sensory features, motion primitives,
and local control laws to move from one distinctive
state to another. Such progressive learning of a
hierarchical representation is called bootstrap
learning. The first step in learning place
recognition eliminates image variability in two
steps: (a) focusing on recognition of distinctive
states defined by the robot's control laws, and (b)
unsupervised learning of clusters of similar sensory
images. The clusters define views associated with
distinctive states, often increasing perceptual
aliasing. The second step eliminates perceptual
aliasing by building a cognitive map and using
history information gathered during exploration to
disambiguate distinctive states. The third step uses
the labeled images for supervised learning of direct
associations from sensory images to distinctive
states. We evaluate the method using a physical
mobile robot in two environments, showing large
amounts of perceptual aliasing and high resulting
recognition rates.},
bib2html_pubtype ={Workshop},
bib2html_rescat ={Foundational Learning},
bib2html_extra_info ={Also see: Benjamin Kuipers and Patrick
Beeson. <a
href="http://personal.traclabs.com/~pbeeson/publications/b2hd-Kuipers-aaai-02.html">
Bootstrap learning for place recognition</a>. AAAI
Conference on Artificial Intelligence, 2002.}
}