OverlapNet: a siamese network for computing LiDAR scan similarity with applications to loop closing and localization View Full Text


Ontology type: schema:ScholarlyArticle      Open Access: True


Article Info

DATE

2021-08-05

AUTHORS

Xieyuanli Chen, Thomas Läbe, Andres Milioto, Timo Röhling, Jens Behley, Cyrill Stachniss

ABSTRACT

Localization and mapping are key capabilities of autonomous systems. In this paper, we propose a modified Siamese network to estimate the similarity between pairs of LiDAR scans recorded by autonomous cars. This can be used to address both, loop closing for SLAM and global localization. Our approach utilizes a deep neural network exploiting different cues generated from LiDAR data. It estimates the similarity between pairs of scans using the concept of image overlap generalized to range images and furthermore provides a relative yaw angle estimate. Based on such predictions, our method is able to detect loop closures in a SLAM system or to globally localize in a given map. For loop closure detection, we use the overlap prediction as the similarity measurement to find loop closure candidates and integrate the candidate selection into an existing SLAM system to improve the mapping performance. For global localization, we propose a novel observation model using the predictions provided by OverlapNet and integrate it into a Monte-Carlo localization framework. We evaluate our approach on multiple datasets collected using different LiDAR scanners in various environments. The experimental results show that our method can effectively detect loop closures surpassing the detection performance of state-of-the-art methods and that it generalizes well to different environments. Furthermore, our method reliably localizes a vehicle in typical urban environments globally using LiDAR data collected in different seasons. More... »

PAGES

61-81

References to SciGraph publications

  • 2016-07-27. Simultaneous Localization and Mapping in SPRINGER HANDBOOK OF ROBOTICS
  • 2007. A High-Accuracy Rotation Estimation Algorithm Based on 1D Phase-Only Correlation in IMAGE ANALYSIS AND RECOGNITION
  • 2017-02-21. Improving SLAM by Exploiting Building Information from Publicly Available Maps and Localization Priors in PFG – JOURNAL OF PHOTOGRAMMETRY, REMOTE SENSING AND GEOINFORMATION SCIENCE
  • Identifiers

    URI

    http://scigraph.springernature.com/pub.10.1007/s10514-021-09999-0

    DOI

    http://dx.doi.org/10.1007/s10514-021-09999-0

    DIMENSIONS

    https://app.dimensions.ai/details/publication/pub.1140215441


    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": "University of Bonn, Bonn, Germany", 
              "id": "http://www.grid.ac/institutes/grid.10388.32", 
              "name": [
                "University of Bonn, Bonn, Germany"
              ], 
              "type": "Organization"
            }, 
            "familyName": "Chen", 
            "givenName": "Xieyuanli", 
            "id": "sg:person.013212367603.97", 
            "sameAs": [
              "https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.013212367603.97"
            ], 
            "type": "Person"
          }, 
          {
            "affiliation": {
              "alternateName": "University of Bonn, Bonn, Germany", 
              "id": "http://www.grid.ac/institutes/grid.10388.32", 
              "name": [
                "University of Bonn, Bonn, Germany"
              ], 
              "type": "Organization"
            }, 
            "familyName": "L\u00e4be", 
            "givenName": "Thomas", 
            "id": "sg:person.013350547723.61", 
            "sameAs": [
              "https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.013350547723.61"
            ], 
            "type": "Person"
          }, 
          {
            "affiliation": {
              "alternateName": "University of Bonn, Bonn, Germany", 
              "id": "http://www.grid.ac/institutes/grid.10388.32", 
              "name": [
                "University of Bonn, Bonn, Germany"
              ], 
              "type": "Organization"
            }, 
            "familyName": "Milioto", 
            "givenName": "Andres", 
            "id": "sg:person.012056355147.01", 
            "sameAs": [
              "https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.012056355147.01"
            ], 
            "type": "Person"
          }, 
          {
            "affiliation": {
              "alternateName": "University of Bonn, Bonn, Germany", 
              "id": "http://www.grid.ac/institutes/grid.10388.32", 
              "name": [
                "University of Bonn, Bonn, Germany"
              ], 
              "type": "Organization"
            }, 
            "familyName": "R\u00f6hling", 
            "givenName": "Timo", 
            "id": "sg:person.014061610176.93", 
            "sameAs": [
              "https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.014061610176.93"
            ], 
            "type": "Person"
          }, 
          {
            "affiliation": {
              "alternateName": "University of Bonn, Bonn, Germany", 
              "id": "http://www.grid.ac/institutes/grid.10388.32", 
              "name": [
                "University of Bonn, Bonn, Germany"
              ], 
              "type": "Organization"
            }, 
            "familyName": "Behley", 
            "givenName": "Jens", 
            "id": "sg:person.01272053671.28", 
            "sameAs": [
              "https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.01272053671.28"
            ], 
            "type": "Person"
          }, 
          {
            "affiliation": {
              "alternateName": "University of Bonn, Bonn, Germany", 
              "id": "http://www.grid.ac/institutes/grid.10388.32", 
              "name": [
                "University of Bonn, Bonn, 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"
          }
        ], 
        "citation": [
          {
            "id": "sg:pub.10.1007/s41064-017-0006-3", 
            "sameAs": [
              "https://app.dimensions.ai/details/publication/pub.1083919644", 
              "https://doi.org/10.1007/s41064-017-0006-3"
            ], 
            "type": "CreativeWork"
          }, 
          {
            "id": "sg:pub.10.1007/978-3-319-32552-1_46", 
            "sameAs": [
              "https://app.dimensions.ai/details/publication/pub.1005252971", 
              "https://doi.org/10.1007/978-3-319-32552-1_46"
            ], 
            "type": "CreativeWork"
          }, 
          {
            "id": "sg:pub.10.1007/978-3-540-74260-9_19", 
            "sameAs": [
              "https://app.dimensions.ai/details/publication/pub.1030030636", 
              "https://doi.org/10.1007/978-3-540-74260-9_19"
            ], 
            "type": "CreativeWork"
          }
        ], 
        "datePublished": "2021-08-05", 
        "datePublishedReg": "2021-08-05", 
        "description": "Localization and mapping are key capabilities of autonomous systems. In this paper, we propose a modified Siamese network to estimate the similarity between pairs of LiDAR scans recorded by autonomous cars. This can be used to address both, loop closing for SLAM and global localization. Our approach utilizes a deep neural network exploiting different cues generated from LiDAR data. It estimates the similarity between pairs of scans using the concept of image overlap generalized to range images and furthermore provides a relative yaw angle estimate. Based on such predictions, our method is able to detect loop closures in a SLAM system or to globally localize in a given map. For loop closure detection, we use the overlap prediction as the similarity measurement to find loop closure candidates and integrate the candidate selection into an existing SLAM system to improve the mapping performance. For global localization, we propose a novel observation model using the predictions provided by OverlapNet and integrate it into a Monte-Carlo localization framework. We evaluate our approach on multiple datasets collected using different LiDAR scanners in various environments. The experimental results show that our method can effectively detect loop closures surpassing the detection performance of state-of-the-art methods and that it generalizes well to different environments. Furthermore, our method reliably localizes a vehicle in typical urban environments globally using LiDAR data collected in different seasons.", 
        "genre": "article", 
        "id": "sg:pub.10.1007/s10514-021-09999-0", 
        "isAccessibleForFree": true, 
        "isPartOf": [
          {
            "id": "sg:journal.1031086", 
            "issn": [
              "0929-5593", 
              "1573-7527"
            ], 
            "name": "Autonomous Robots", 
            "publisher": "Springer Nature", 
            "type": "Periodical"
          }, 
          {
            "issueNumber": "1", 
            "type": "PublicationIssue"
          }, 
          {
            "type": "PublicationVolume", 
            "volumeNumber": "46"
          }
        ], 
        "keywords": [
          "Siamese network", 
          "SLAM system", 
          "global localization", 
          "loop closure", 
          "loop closure candidates", 
          "deep neural networks", 
          "loop closure detection", 
          "Monte-Carlo localization framework", 
          "novel observation model", 
          "LiDAR data", 
          "closure detection", 
          "autonomous cars", 
          "localization framework", 
          "neural network", 
          "art methods", 
          "similarity measurement", 
          "autonomous systems", 
          "image overlap", 
          "key capabilities", 
          "LiDAR scanner", 
          "multiple datasets", 
          "LiDAR scans", 
          "detection performance", 
          "mapping performance", 
          "typical urban environment", 
          "network", 
          "observation model", 
          "experimental results", 
          "pairs of scans", 
          "environment", 
          "SLAM", 
          "such predictions", 
          "different environments", 
          "urban environment", 
          "candidate selection", 
          "dataset", 
          "system", 
          "performance", 
          "images", 
          "different cues", 
          "prediction", 
          "angle estimates", 
          "framework", 
          "capability", 
          "method", 
          "similarity", 
          "cars", 
          "vehicles", 
          "localization", 
          "data", 
          "applications", 
          "detection", 
          "maps", 
          "scanner", 
          "mapping", 
          "concept", 
          "selection", 
          "model", 
          "pairs", 
          "cues", 
          "state", 
          "results", 
          "scans", 
          "loop", 
          "candidates", 
          "overlap", 
          "closing", 
          "estimates", 
          "measurements", 
          "approach", 
          "different seasons", 
          "closure", 
          "season", 
          "paper"
        ], 
        "name": "OverlapNet: a siamese network for computing LiDAR scan similarity with applications to loop closing and localization", 
        "pagination": "61-81", 
        "productId": [
          {
            "name": "dimensions_id", 
            "type": "PropertyValue", 
            "value": [
              "pub.1140215441"
            ]
          }, 
          {
            "name": "doi", 
            "type": "PropertyValue", 
            "value": [
              "10.1007/s10514-021-09999-0"
            ]
          }
        ], 
        "sameAs": [
          "https://doi.org/10.1007/s10514-021-09999-0", 
          "https://app.dimensions.ai/details/publication/pub.1140215441"
        ], 
        "sdDataset": "articles", 
        "sdDatePublished": "2022-11-24T21:07", 
        "sdLicense": "https://scigraph.springernature.com/explorer/license/", 
        "sdPublisher": {
          "name": "Springer Nature - SN SciGraph project", 
          "type": "Organization"
        }, 
        "sdSource": "s3://com-springernature-scigraph/baseset/20221124/entities/gbq_results/article/article_911.jsonl", 
        "type": "ScholarlyArticle", 
        "url": "https://doi.org/10.1007/s10514-021-09999-0"
      }
    ]
     

    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/s10514-021-09999-0'

    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/s10514-021-09999-0'

    Turtle is a human-readable linked data format.

    curl -H 'Accept: text/turtle' 'https://scigraph.springernature.com/pub.10.1007/s10514-021-09999-0'

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

    curl -H 'Accept: application/rdf+xml' 'https://scigraph.springernature.com/pub.10.1007/s10514-021-09999-0'


     

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

    178 TRIPLES      21 PREDICATES      101 URIs      90 LITERALS      6 BLANK NODES

    Subject Predicate Object
    1 sg:pub.10.1007/s10514-021-09999-0 schema:about anzsrc-for:08
    2 anzsrc-for:0801
    3 schema:author Neb36abb4cb3842dc89814591f3d86241
    4 schema:citation sg:pub.10.1007/978-3-319-32552-1_46
    5 sg:pub.10.1007/978-3-540-74260-9_19
    6 sg:pub.10.1007/s41064-017-0006-3
    7 schema:datePublished 2021-08-05
    8 schema:datePublishedReg 2021-08-05
    9 schema:description Localization and mapping are key capabilities of autonomous systems. In this paper, we propose a modified Siamese network to estimate the similarity between pairs of LiDAR scans recorded by autonomous cars. This can be used to address both, loop closing for SLAM and global localization. Our approach utilizes a deep neural network exploiting different cues generated from LiDAR data. It estimates the similarity between pairs of scans using the concept of image overlap generalized to range images and furthermore provides a relative yaw angle estimate. Based on such predictions, our method is able to detect loop closures in a SLAM system or to globally localize in a given map. For loop closure detection, we use the overlap prediction as the similarity measurement to find loop closure candidates and integrate the candidate selection into an existing SLAM system to improve the mapping performance. For global localization, we propose a novel observation model using the predictions provided by OverlapNet and integrate it into a Monte-Carlo localization framework. We evaluate our approach on multiple datasets collected using different LiDAR scanners in various environments. The experimental results show that our method can effectively detect loop closures surpassing the detection performance of state-of-the-art methods and that it generalizes well to different environments. Furthermore, our method reliably localizes a vehicle in typical urban environments globally using LiDAR data collected in different seasons.
    10 schema:genre article
    11 schema:isAccessibleForFree true
    12 schema:isPartOf N1592159d387b44c39f28c26d6cdcdbce
    13 Ndfb75c4665454682baeb359c917547e7
    14 sg:journal.1031086
    15 schema:keywords LiDAR data
    16 LiDAR scanner
    17 LiDAR scans
    18 Monte-Carlo localization framework
    19 SLAM
    20 SLAM system
    21 Siamese network
    22 angle estimates
    23 applications
    24 approach
    25 art methods
    26 autonomous cars
    27 autonomous systems
    28 candidate selection
    29 candidates
    30 capability
    31 cars
    32 closing
    33 closure
    34 closure detection
    35 concept
    36 cues
    37 data
    38 dataset
    39 deep neural networks
    40 detection
    41 detection performance
    42 different cues
    43 different environments
    44 different seasons
    45 environment
    46 estimates
    47 experimental results
    48 framework
    49 global localization
    50 image overlap
    51 images
    52 key capabilities
    53 localization
    54 localization framework
    55 loop
    56 loop closure
    57 loop closure candidates
    58 loop closure detection
    59 mapping
    60 mapping performance
    61 maps
    62 measurements
    63 method
    64 model
    65 multiple datasets
    66 network
    67 neural network
    68 novel observation model
    69 observation model
    70 overlap
    71 pairs
    72 pairs of scans
    73 paper
    74 performance
    75 prediction
    76 results
    77 scanner
    78 scans
    79 season
    80 selection
    81 similarity
    82 similarity measurement
    83 state
    84 such predictions
    85 system
    86 typical urban environment
    87 urban environment
    88 vehicles
    89 schema:name OverlapNet: a siamese network for computing LiDAR scan similarity with applications to loop closing and localization
    90 schema:pagination 61-81
    91 schema:productId N14d8db88fb24488ebab69ec227234c53
    92 N2c99c3016d4c4b978c5d0ed719cee04c
    93 schema:sameAs https://app.dimensions.ai/details/publication/pub.1140215441
    94 https://doi.org/10.1007/s10514-021-09999-0
    95 schema:sdDatePublished 2022-11-24T21:07
    96 schema:sdLicense https://scigraph.springernature.com/explorer/license/
    97 schema:sdPublisher Nef88c611c8f046e89f2ae5c5ac5522ed
    98 schema:url https://doi.org/10.1007/s10514-021-09999-0
    99 sgo:license sg:explorer/license/
    100 sgo:sdDataset articles
    101 rdf:type schema:ScholarlyArticle
    102 N14d8db88fb24488ebab69ec227234c53 schema:name doi
    103 schema:value 10.1007/s10514-021-09999-0
    104 rdf:type schema:PropertyValue
    105 N1592159d387b44c39f28c26d6cdcdbce schema:volumeNumber 46
    106 rdf:type schema:PublicationVolume
    107 N1efd134433f24aaa81705296022a5faa rdf:first sg:person.013350547723.61
    108 rdf:rest Nc8e54e84f65743eca1eefdee2da961c6
    109 N2c99c3016d4c4b978c5d0ed719cee04c schema:name dimensions_id
    110 schema:value pub.1140215441
    111 rdf:type schema:PropertyValue
    112 N6b0e414dd87a4df6a4bcdc9e07b1243e rdf:first sg:person.015152144445.37
    113 rdf:rest rdf:nil
    114 N8786d52e7a794a4b9f30df5574d754c2 rdf:first sg:person.01272053671.28
    115 rdf:rest N6b0e414dd87a4df6a4bcdc9e07b1243e
    116 Nc8e54e84f65743eca1eefdee2da961c6 rdf:first sg:person.012056355147.01
    117 rdf:rest Nf52d532e144b4084904e37b14aa42128
    118 Ndfb75c4665454682baeb359c917547e7 schema:issueNumber 1
    119 rdf:type schema:PublicationIssue
    120 Neb36abb4cb3842dc89814591f3d86241 rdf:first sg:person.013212367603.97
    121 rdf:rest N1efd134433f24aaa81705296022a5faa
    122 Nef88c611c8f046e89f2ae5c5ac5522ed schema:name Springer Nature - SN SciGraph project
    123 rdf:type schema:Organization
    124 Nf52d532e144b4084904e37b14aa42128 rdf:first sg:person.014061610176.93
    125 rdf:rest N8786d52e7a794a4b9f30df5574d754c2
    126 anzsrc-for:08 schema:inDefinedTermSet anzsrc-for:
    127 schema:name Information and Computing Sciences
    128 rdf:type schema:DefinedTerm
    129 anzsrc-for:0801 schema:inDefinedTermSet anzsrc-for:
    130 schema:name Artificial Intelligence and Image Processing
    131 rdf:type schema:DefinedTerm
    132 sg:journal.1031086 schema:issn 0929-5593
    133 1573-7527
    134 schema:name Autonomous Robots
    135 schema:publisher Springer Nature
    136 rdf:type schema:Periodical
    137 sg:person.012056355147.01 schema:affiliation grid-institutes:grid.10388.32
    138 schema:familyName Milioto
    139 schema:givenName Andres
    140 schema:sameAs https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.012056355147.01
    141 rdf:type schema:Person
    142 sg:person.01272053671.28 schema:affiliation grid-institutes:grid.10388.32
    143 schema:familyName Behley
    144 schema:givenName Jens
    145 schema:sameAs https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.01272053671.28
    146 rdf:type schema:Person
    147 sg:person.013212367603.97 schema:affiliation grid-institutes:grid.10388.32
    148 schema:familyName Chen
    149 schema:givenName Xieyuanli
    150 schema:sameAs https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.013212367603.97
    151 rdf:type schema:Person
    152 sg:person.013350547723.61 schema:affiliation grid-institutes:grid.10388.32
    153 schema:familyName Läbe
    154 schema:givenName Thomas
    155 schema:sameAs https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.013350547723.61
    156 rdf:type schema:Person
    157 sg:person.014061610176.93 schema:affiliation grid-institutes:grid.10388.32
    158 schema:familyName Röhling
    159 schema:givenName Timo
    160 schema:sameAs https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.014061610176.93
    161 rdf:type schema:Person
    162 sg:person.015152144445.37 schema:affiliation grid-institutes:grid.10388.32
    163 schema:familyName Stachniss
    164 schema:givenName Cyrill
    165 schema:sameAs https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.015152144445.37
    166 rdf:type schema:Person
    167 sg:pub.10.1007/978-3-319-32552-1_46 schema:sameAs https://app.dimensions.ai/details/publication/pub.1005252971
    168 https://doi.org/10.1007/978-3-319-32552-1_46
    169 rdf:type schema:CreativeWork
    170 sg:pub.10.1007/978-3-540-74260-9_19 schema:sameAs https://app.dimensions.ai/details/publication/pub.1030030636
    171 https://doi.org/10.1007/978-3-540-74260-9_19
    172 rdf:type schema:CreativeWork
    173 sg:pub.10.1007/s41064-017-0006-3 schema:sameAs https://app.dimensions.ai/details/publication/pub.1083919644
    174 https://doi.org/10.1007/s41064-017-0006-3
    175 rdf:type schema:CreativeWork
    176 grid-institutes:grid.10388.32 schema:alternateName University of Bonn, Bonn, Germany
    177 schema:name University of Bonn, Bonn, Germany
    178 rdf:type schema:Organization
     




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


    ...