Learning Landmark Selection Policies for Mapping Unknown Environments View Full Text


Ontology type: schema:Chapter      Open Access: True


Chapter Info

DATE

2011

AUTHORS

Hauke Strasdat , Cyrill Stachniss , Wolfram Burgard

ABSTRACT

In general, a mobile robot that operates in unknown environments has to maintain a map and has to determine its own location given themap. This introduces significant computational and memory constraints for most autonomous systems, especially for lightweight robots such as humanoids or flying vehicles. In this paper, we present a universal approach for learning a landmark selection policy that allows a robot to discard landmarks that are not valuable for its current navigation task. This enables the robot to reduce the computational burden and to carry out its task more efficiently by maintaining only the important landmarks. Our approach applies an unscented Kalman filter for addressing the simultaneous localization and mapping problem and uses Monte-Carlo reinforcement learning to obtain the selection policy. In addition to that, we present a technique to compress learned policies without introducing a performance loss. In this way, our approach becomes applicable on systems with constrained memory resources. Based on real world and simulation experiments, we show that the learned policies allow for efficient robot navigation and outperform handcrafted strategies.We furthermore demonstrate that the learned policies are not only usable in a specific scenario but can also be generalized towards environments with varying properties. More... »

PAGES

483-499

Identifiers

URI

http://scigraph.springernature.com/pub.10.1007/978-3-642-19457-3_29

DOI

http://dx.doi.org/10.1007/978-3-642-19457-3_29

DIMENSIONS

https://app.dimensions.ai/details/publication/pub.1045526053


Indexing Status Check whether this publication has been indexed by Scopus and Web Of Science using the SN Indexing Status Tool
Incoming Citations Browse incoming citations for this publication using opencitations.net

JSON-LD is the canonical representation for SciGraph data.

TIP: You can open this SciGraph record using an external JSON-LD service: JSON-LD Playground Google SDTT

[
  {
    "@context": "https://springernature.github.io/scigraph/jsonld/sgcontext.json", 
    "about": [
      {
        "id": "http://purl.org/au-research/vocabulary/anzsrc-for/2008/08", 
        "inDefinedTermSet": "http://purl.org/au-research/vocabulary/anzsrc-for/2008/", 
        "name": "Information and Computing Sciences", 
        "type": "DefinedTerm"
      }, 
      {
        "id": "http://purl.org/au-research/vocabulary/anzsrc-for/2008/0801", 
        "inDefinedTermSet": "http://purl.org/au-research/vocabulary/anzsrc-for/2008/", 
        "name": "Artificial Intelligence and Image Processing", 
        "type": "DefinedTerm"
      }
    ], 
    "author": [
      {
        "affiliation": {
          "alternateName": "Dept.\u00a0of Computing, Imperial College London, SW7 2AZ, UK", 
          "id": "http://www.grid.ac/institutes/grid.7445.2", 
          "name": [
            "Dept.\u00a0of Computing, Imperial College London, SW7 2AZ, UK"
          ], 
          "type": "Organization"
        }, 
        "familyName": "Strasdat", 
        "givenName": "Hauke", 
        "id": "sg:person.010640440533.15", 
        "sameAs": [
          "https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.010640440533.15"
        ], 
        "type": "Person"
      }, 
      {
        "affiliation": {
          "alternateName": "Dept.\u00a0of Computer Science, University of Freiburg, D-79110, Freiburg, Germany", 
          "id": "http://www.grid.ac/institutes/grid.5963.9", 
          "name": [
            "Dept.\u00a0of Computer Science, University of Freiburg, D-79110, Freiburg, Germany"
          ], 
          "type": "Organization"
        }, 
        "familyName": "Stachniss", 
        "givenName": "Cyrill", 
        "id": "sg:person.015152144445.37", 
        "sameAs": [
          "https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.015152144445.37"
        ], 
        "type": "Person"
      }, 
      {
        "affiliation": {
          "alternateName": "Dept.\u00a0of Computer Science, University of Freiburg, D-79110, Freiburg, Germany", 
          "id": "http://www.grid.ac/institutes/grid.5963.9", 
          "name": [
            "Dept.\u00a0of Computer Science, University of Freiburg, D-79110, Freiburg, Germany"
          ], 
          "type": "Organization"
        }, 
        "familyName": "Burgard", 
        "givenName": "Wolfram", 
        "id": "sg:person.014270043511.25", 
        "sameAs": [
          "https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.014270043511.25"
        ], 
        "type": "Person"
      }
    ], 
    "datePublished": "2011", 
    "datePublishedReg": "2011-01-01", 
    "description": "In general, a mobile robot that operates in unknown environments has to maintain a map and has to determine its own location given themap. This introduces significant computational and memory constraints for most autonomous systems, especially for lightweight robots such as humanoids or flying vehicles. In this paper, we present a universal approach for learning a landmark selection policy that allows a robot to discard landmarks that are not valuable for its current navigation task. This enables the robot to reduce the computational burden and to carry out its task more efficiently by maintaining only the important landmarks. Our approach applies an unscented Kalman filter for addressing the simultaneous localization and mapping problem and uses Monte-Carlo reinforcement learning to obtain the selection policy. In addition to that, we present a technique to compress learned policies without introducing a performance loss. In this way, our approach becomes applicable on systems with constrained memory resources. Based on real world and simulation experiments, we show that the learned policies allow for efficient robot navigation and outperform handcrafted strategies.We furthermore demonstrate that the learned policies are not only usable in a specific scenario but can also be generalized towards environments with varying properties.", 
    "editor": [
      {
        "familyName": "Pradalier", 
        "givenName": "C\u00e9dric", 
        "type": "Person"
      }, 
      {
        "familyName": "Siegwart", 
        "givenName": "Roland", 
        "type": "Person"
      }, 
      {
        "familyName": "Hirzinger", 
        "givenName": "Gerhard", 
        "type": "Person"
      }
    ], 
    "genre": "chapter", 
    "id": "sg:pub.10.1007/978-3-642-19457-3_29", 
    "isAccessibleForFree": true, 
    "isPartOf": {
      "isbn": [
        "978-3-642-19456-6", 
        "978-3-642-19457-3"
      ], 
      "name": "Robotics Research", 
      "type": "Book"
    }, 
    "keywords": [
      "unknown environment", 
      "selection policy", 
      "Monte Carlo reinforcement", 
      "efficient robot navigation", 
      "mapping unknown environments", 
      "most autonomous systems", 
      "robot navigation", 
      "mobile robot", 
      "significant computational", 
      "memory constraints", 
      "simultaneous localization", 
      "memory resources", 
      "navigation task", 
      "lightweight robot", 
      "mapping problem", 
      "autonomous systems", 
      "own location", 
      "robot", 
      "real world", 
      "computational burden", 
      "specific scenarios", 
      "simulation experiments", 
      "performance loss", 
      "task", 
      "Kalman filter", 
      "unscented Kalman filter", 
      "environment", 
      "humanoid", 
      "navigation", 
      "computational", 
      "landmarks", 
      "universal approach", 
      "system", 
      "scenarios", 
      "constraints", 
      "themap", 
      "resources", 
      "vehicles", 
      "maps", 
      "technique", 
      "filter", 
      "way", 
      "policy", 
      "important landmark", 
      "world", 
      "experiments", 
      "location", 
      "localization", 
      "strategies", 
      "reinforcement", 
      "burden", 
      "addition", 
      "loss", 
      "properties", 
      "approach", 
      "paper", 
      "problem"
    ], 
    "name": "Learning Landmark Selection Policies for Mapping Unknown Environments", 
    "pagination": "483-499", 
    "productId": [
      {
        "name": "dimensions_id", 
        "type": "PropertyValue", 
        "value": [
          "pub.1045526053"
        ]
      }, 
      {
        "name": "doi", 
        "type": "PropertyValue", 
        "value": [
          "10.1007/978-3-642-19457-3_29"
        ]
      }
    ], 
    "publisher": {
      "name": "Springer Nature", 
      "type": "Organisation"
    }, 
    "sameAs": [
      "https://doi.org/10.1007/978-3-642-19457-3_29", 
      "https://app.dimensions.ai/details/publication/pub.1045526053"
    ], 
    "sdDataset": "chapters", 
    "sdDatePublished": "2022-12-01T06:51", 
    "sdLicense": "https://scigraph.springernature.com/explorer/license/", 
    "sdPublisher": {
      "name": "Springer Nature - SN SciGraph project", 
      "type": "Organization"
    }, 
    "sdSource": "s3://com-springernature-scigraph/baseset/20221201/entities/gbq_results/chapter/chapter_333.jsonl", 
    "type": "Chapter", 
    "url": "https://doi.org/10.1007/978-3-642-19457-3_29"
  }
]
 

Download the RDF metadata as:  json-ld nt turtle xml License info

HOW TO GET THIS DATA PROGRAMMATICALLY:

JSON-LD is a popular format for linked data which is fully compatible with JSON.

curl -H 'Accept: application/ld+json' 'https://scigraph.springernature.com/pub.10.1007/978-3-642-19457-3_29'

N-Triples is a line-based linked data format ideal for batch operations.

curl -H 'Accept: application/n-triples' 'https://scigraph.springernature.com/pub.10.1007/978-3-642-19457-3_29'

Turtle is a human-readable linked data format.

curl -H 'Accept: text/turtle' 'https://scigraph.springernature.com/pub.10.1007/978-3-642-19457-3_29'

RDF/XML is a standard XML format for linked data.

curl -H 'Accept: application/rdf+xml' 'https://scigraph.springernature.com/pub.10.1007/978-3-642-19457-3_29'


 

This table displays all metadata directly associated to this object as RDF triples.

143 TRIPLES      22 PREDICATES      82 URIs      75 LITERALS      7 BLANK NODES

Subject Predicate Object
1 sg:pub.10.1007/978-3-642-19457-3_29 schema:about anzsrc-for:08
2 anzsrc-for:0801
3 schema:author N30195665d7ff44b0bbd950d19a8d6814
4 schema:datePublished 2011
5 schema:datePublishedReg 2011-01-01
6 schema:description In general, a mobile robot that operates in unknown environments has to maintain a map and has to determine its own location given themap. This introduces significant computational and memory constraints for most autonomous systems, especially for lightweight robots such as humanoids or flying vehicles. In this paper, we present a universal approach for learning a landmark selection policy that allows a robot to discard landmarks that are not valuable for its current navigation task. This enables the robot to reduce the computational burden and to carry out its task more efficiently by maintaining only the important landmarks. Our approach applies an unscented Kalman filter for addressing the simultaneous localization and mapping problem and uses Monte-Carlo reinforcement learning to obtain the selection policy. In addition to that, we present a technique to compress learned policies without introducing a performance loss. In this way, our approach becomes applicable on systems with constrained memory resources. Based on real world and simulation experiments, we show that the learned policies allow for efficient robot navigation and outperform handcrafted strategies.We furthermore demonstrate that the learned policies are not only usable in a specific scenario but can also be generalized towards environments with varying properties.
7 schema:editor N33afa8b06d5f4c4fb560185bcabbb832
8 schema:genre chapter
9 schema:isAccessibleForFree true
10 schema:isPartOf N09c518ced5ac4d81929d30c663855600
11 schema:keywords Kalman filter
12 Monte Carlo reinforcement
13 addition
14 approach
15 autonomous systems
16 burden
17 computational
18 computational burden
19 constraints
20 efficient robot navigation
21 environment
22 experiments
23 filter
24 humanoid
25 important landmark
26 landmarks
27 lightweight robot
28 localization
29 location
30 loss
31 mapping problem
32 mapping unknown environments
33 maps
34 memory constraints
35 memory resources
36 mobile robot
37 most autonomous systems
38 navigation
39 navigation task
40 own location
41 paper
42 performance loss
43 policy
44 problem
45 properties
46 real world
47 reinforcement
48 resources
49 robot
50 robot navigation
51 scenarios
52 selection policy
53 significant computational
54 simulation experiments
55 simultaneous localization
56 specific scenarios
57 strategies
58 system
59 task
60 technique
61 themap
62 universal approach
63 unknown environment
64 unscented Kalman filter
65 vehicles
66 way
67 world
68 schema:name Learning Landmark Selection Policies for Mapping Unknown Environments
69 schema:pagination 483-499
70 schema:productId N516d3761e4434dd385c4abeaa8369bf4
71 N8afc47ddb5b0442fb4cc27d762cdd186
72 schema:publisher Nb62b63dd1ed842e9a43942a77ea142b1
73 schema:sameAs https://app.dimensions.ai/details/publication/pub.1045526053
74 https://doi.org/10.1007/978-3-642-19457-3_29
75 schema:sdDatePublished 2022-12-01T06:51
76 schema:sdLicense https://scigraph.springernature.com/explorer/license/
77 schema:sdPublisher Na295247744a14b608b03152991b32094
78 schema:url https://doi.org/10.1007/978-3-642-19457-3_29
79 sgo:license sg:explorer/license/
80 sgo:sdDataset chapters
81 rdf:type schema:Chapter
82 N09c518ced5ac4d81929d30c663855600 schema:isbn 978-3-642-19456-6
83 978-3-642-19457-3
84 schema:name Robotics Research
85 rdf:type schema:Book
86 N2668b18de7834887b863a8115657abe9 schema:familyName Pradalier
87 schema:givenName Cédric
88 rdf:type schema:Person
89 N30195665d7ff44b0bbd950d19a8d6814 rdf:first sg:person.010640440533.15
90 rdf:rest Na3b1810c7b734475a58b818ae6b83b19
91 N33afa8b06d5f4c4fb560185bcabbb832 rdf:first N2668b18de7834887b863a8115657abe9
92 rdf:rest N7dd09890e80448b4ac5010ead9007b83
93 N516d3761e4434dd385c4abeaa8369bf4 schema:name dimensions_id
94 schema:value pub.1045526053
95 rdf:type schema:PropertyValue
96 N568376e6cd5d4cc8a43b03c4ab0cd8fe schema:familyName Siegwart
97 schema:givenName Roland
98 rdf:type schema:Person
99 N7dd09890e80448b4ac5010ead9007b83 rdf:first N568376e6cd5d4cc8a43b03c4ab0cd8fe
100 rdf:rest Na92cc24687464aff8adf21e0b271bbb9
101 N8afc47ddb5b0442fb4cc27d762cdd186 schema:name doi
102 schema:value 10.1007/978-3-642-19457-3_29
103 rdf:type schema:PropertyValue
104 N96b032f5424f4096b039433a953e1909 schema:familyName Hirzinger
105 schema:givenName Gerhard
106 rdf:type schema:Person
107 Na295247744a14b608b03152991b32094 schema:name Springer Nature - SN SciGraph project
108 rdf:type schema:Organization
109 Na3b1810c7b734475a58b818ae6b83b19 rdf:first sg:person.015152144445.37
110 rdf:rest Nb097930395f04ea0a4f34b0ca9417316
111 Na92cc24687464aff8adf21e0b271bbb9 rdf:first N96b032f5424f4096b039433a953e1909
112 rdf:rest rdf:nil
113 Nb097930395f04ea0a4f34b0ca9417316 rdf:first sg:person.014270043511.25
114 rdf:rest rdf:nil
115 Nb62b63dd1ed842e9a43942a77ea142b1 schema:name Springer Nature
116 rdf:type schema:Organisation
117 anzsrc-for:08 schema:inDefinedTermSet anzsrc-for:
118 schema:name Information and Computing Sciences
119 rdf:type schema:DefinedTerm
120 anzsrc-for:0801 schema:inDefinedTermSet anzsrc-for:
121 schema:name Artificial Intelligence and Image Processing
122 rdf:type schema:DefinedTerm
123 sg:person.010640440533.15 schema:affiliation grid-institutes:grid.7445.2
124 schema:familyName Strasdat
125 schema:givenName Hauke
126 schema:sameAs https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.010640440533.15
127 rdf:type schema:Person
128 sg:person.014270043511.25 schema:affiliation grid-institutes:grid.5963.9
129 schema:familyName Burgard
130 schema:givenName Wolfram
131 schema:sameAs https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.014270043511.25
132 rdf:type schema:Person
133 sg:person.015152144445.37 schema:affiliation grid-institutes:grid.5963.9
134 schema:familyName Stachniss
135 schema:givenName Cyrill
136 schema:sameAs https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.015152144445.37
137 rdf:type schema:Person
138 grid-institutes:grid.5963.9 schema:alternateName Dept. of Computer Science, University of Freiburg, D-79110, Freiburg, Germany
139 schema:name Dept. of Computer Science, University of Freiburg, D-79110, Freiburg, Germany
140 rdf:type schema:Organization
141 grid-institutes:grid.7445.2 schema:alternateName Dept. of Computing, Imperial College London, SW7 2AZ, UK
142 schema:name Dept. of Computing, Imperial College London, SW7 2AZ, UK
143 rdf:type schema:Organization
 




Preview window. Press ESC to close (or click here)


...