A Category-Level 3D Object Dataset: Putting the Kinect to Work View Full Text


Ontology type: schema:Chapter     


Chapter Info

DATE

2013

AUTHORS

Allison Janoch , Sergey Karayev , Yangqing Jia , Jonathan T. Barron , Mario Fritz , Kate Saenko , Trevor Darrell

ABSTRACT

The recent proliferation of the Microsoft Kinect, a cheap but quality depth sensor, has brought the need for a challenging category-level 3D object detection dataset to the forefront. Such a dataset can be used for object recognition in a spirit usually reserved for the large collections of intensity images typically collected from the Internet. Here, we will review current 3D datasets and find them lacking in variation of scene, category, instance, and viewpoint. The Berkeley 3D Object Dataset (B3DO), which contains color and depth image pairs gathered in read domestic and office environments will be presented. Baseline object recognition performance in a PASCAL VOC-style detection task is established, and two ways that inferred world size of the object van be used to improve detection are suggested. In an effort to make more significant performance progress, the problem of extracting useful features from range images is addressed. There has been much success in using the histogram of oriented gradients (HOG) as a global descriptor for object detection in intensity images. There are also many proposed descriptors designed specifically for depth data (spin images, shape context, etc.), but these are often focused on the local, not global descriptor paradigm. We explore the failures of gradient-based descriptors when applied to depth, and propose that the proper global descriptor in the realm of 3D should be based on curvature, not gradients. More... »

PAGES

141-165

Identifiers

URI

http://scigraph.springernature.com/pub.10.1007/978-1-4471-4640-7_8

DOI

http://dx.doi.org/10.1007/978-1-4471-4640-7_8

DIMENSIONS

https://app.dimensions.ai/details/publication/pub.1030895778


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 California at Berkeley, Berkeley, CA, USA", 
          "id": "http://www.grid.ac/institutes/grid.47840.3f", 
          "name": [
            "University of California at Berkeley, Berkeley, CA, USA"
          ], 
          "type": "Organization"
        }, 
        "familyName": "Janoch", 
        "givenName": "Allison", 
        "id": "sg:person.014576140257.12", 
        "sameAs": [
          "https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.014576140257.12"
        ], 
        "type": "Person"
      }, 
      {
        "affiliation": {
          "alternateName": "University of California at Berkeley, Berkeley, CA, USA", 
          "id": "http://www.grid.ac/institutes/grid.47840.3f", 
          "name": [
            "University of California at Berkeley, Berkeley, CA, USA"
          ], 
          "type": "Organization"
        }, 
        "familyName": "Karayev", 
        "givenName": "Sergey", 
        "id": "sg:person.016171101257.73", 
        "sameAs": [
          "https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.016171101257.73"
        ], 
        "type": "Person"
      }, 
      {
        "affiliation": {
          "alternateName": "University of California at Berkeley, Berkeley, CA, USA", 
          "id": "http://www.grid.ac/institutes/grid.47840.3f", 
          "name": [
            "University of California at Berkeley, Berkeley, CA, USA"
          ], 
          "type": "Organization"
        }, 
        "familyName": "Jia", 
        "givenName": "Yangqing", 
        "id": "sg:person.013220623467.72", 
        "sameAs": [
          "https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.013220623467.72"
        ], 
        "type": "Person"
      }, 
      {
        "affiliation": {
          "alternateName": "University of California at Berkeley, Berkeley, CA, USA", 
          "id": "http://www.grid.ac/institutes/grid.47840.3f", 
          "name": [
            "University of California at Berkeley, Berkeley, CA, USA"
          ], 
          "type": "Organization"
        }, 
        "familyName": "Barron", 
        "givenName": "Jonathan T.", 
        "id": "sg:person.01302212713.98", 
        "sameAs": [
          "https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.01302212713.98"
        ], 
        "type": "Person"
      }, 
      {
        "affiliation": {
          "alternateName": "Max Plank Institute for Informatics, Campus E1.4, 66123, Saarbr\u00fccken, Germany", 
          "id": "http://www.grid.ac/institutes/None", 
          "name": [
            "Max Plank Institute for Informatics, Campus E1.4, 66123, Saarbr\u00fccken, Germany"
          ], 
          "type": "Organization"
        }, 
        "familyName": "Fritz", 
        "givenName": "Mario", 
        "id": "sg:person.013361072755.17", 
        "sameAs": [
          "https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.013361072755.17"
        ], 
        "type": "Person"
      }, 
      {
        "affiliation": {
          "alternateName": "University of California at Berkeley, Berkeley, CA, USA", 
          "id": "http://www.grid.ac/institutes/grid.47840.3f", 
          "name": [
            "University of California at Berkeley, Berkeley, CA, USA"
          ], 
          "type": "Organization"
        }, 
        "familyName": "Saenko", 
        "givenName": "Kate", 
        "id": "sg:person.0617252060.44", 
        "sameAs": [
          "https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.0617252060.44"
        ], 
        "type": "Person"
      }, 
      {
        "affiliation": {
          "alternateName": "University of California at Berkeley, Berkeley, CA, USA", 
          "id": "http://www.grid.ac/institutes/grid.47840.3f", 
          "name": [
            "University of California at Berkeley, Berkeley, CA, USA"
          ], 
          "type": "Organization"
        }, 
        "familyName": "Darrell", 
        "givenName": "Trevor", 
        "id": "sg:person.01001613660.25", 
        "sameAs": [
          "https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.01001613660.25"
        ], 
        "type": "Person"
      }
    ], 
    "datePublished": "2013", 
    "datePublishedReg": "2013-01-01", 
    "description": "The recent proliferation of the Microsoft Kinect, a cheap but quality depth sensor, has brought the need for a challenging category-level 3D object detection dataset to the forefront. Such a dataset can be used for object recognition in a spirit usually reserved for the large collections of intensity images typically collected from the Internet. Here, we will review current 3D datasets and find them lacking in variation of scene, category, instance, and viewpoint. The Berkeley 3D Object Dataset (B3DO), which contains color and depth image pairs gathered in read domestic and office environments will be presented. Baseline object recognition performance in a PASCAL VOC-style detection task is established, and two ways that inferred world size of the object van be used to improve detection are suggested. In an effort to make more significant performance progress, the problem of extracting useful features from range images is addressed. There has been much success in using the histogram of oriented gradients (HOG) as a global descriptor for object detection in intensity images. There are also many proposed descriptors designed specifically for depth data (spin images, shape context, etc.), but these are often focused on the local, not global descriptor paradigm. We explore the failures of gradient-based descriptors when applied to depth, and propose that the proper global descriptor in the realm of 3D should be based on curvature, not gradients.", 
    "editor": [
      {
        "familyName": "Fossati", 
        "givenName": "Andrea", 
        "type": "Person"
      }, 
      {
        "familyName": "Gall", 
        "givenName": "Juergen", 
        "type": "Person"
      }, 
      {
        "familyName": "Grabner", 
        "givenName": "Helmut", 
        "type": "Person"
      }, 
      {
        "familyName": "Ren", 
        "givenName": "Xiaofeng", 
        "type": "Person"
      }, 
      {
        "familyName": "Konolige", 
        "givenName": "Kurt", 
        "type": "Person"
      }
    ], 
    "genre": "chapter", 
    "id": "sg:pub.10.1007/978-1-4471-4640-7_8", 
    "isAccessibleForFree": false, 
    "isPartOf": {
      "isbn": [
        "978-1-4471-4639-1", 
        "978-1-4471-4640-7"
      ], 
      "name": "Consumer Depth Cameras for Computer Vision", 
      "type": "Book"
    }, 
    "keywords": [
      "object detection", 
      "object dataset", 
      "global descriptors", 
      "depth image pairs", 
      "intensity images", 
      "gradient-based descriptors", 
      "variations of scenes", 
      "object recognition performance", 
      "Microsoft Kinect", 
      "depth sensor", 
      "image pairs", 
      "object recognition", 
      "world size", 
      "range images", 
      "recognition performance", 
      "depth data", 
      "useful features", 
      "detection task", 
      "office environment", 
      "large collection", 
      "dataset", 
      "Kinect", 
      "recent proliferation", 
      "descriptors", 
      "images", 
      "Internet", 
      "detection", 
      "scene", 
      "task", 
      "histogram", 
      "recognition", 
      "paradigm", 
      "sensors", 
      "instances", 
      "performance progress", 
      "environment", 
      "performance", 
      "collection", 
      "features", 
      "viewpoint", 
      "way", 
      "need", 
      "data", 
      "efforts", 
      "success", 
      "color", 
      "realm", 
      "categories", 
      "forefront", 
      "progress", 
      "pairs", 
      "size", 
      "failure", 
      "gradient", 
      "depth", 
      "Van", 
      "spirit", 
      "variation", 
      "curvature", 
      "proliferation", 
      "problem"
    ], 
    "name": "A Category-Level 3D Object Dataset: Putting the Kinect to Work", 
    "pagination": "141-165", 
    "productId": [
      {
        "name": "dimensions_id", 
        "type": "PropertyValue", 
        "value": [
          "pub.1030895778"
        ]
      }, 
      {
        "name": "doi", 
        "type": "PropertyValue", 
        "value": [
          "10.1007/978-1-4471-4640-7_8"
        ]
      }
    ], 
    "publisher": {
      "name": "Springer Nature", 
      "type": "Organisation"
    }, 
    "sameAs": [
      "https://doi.org/10.1007/978-1-4471-4640-7_8", 
      "https://app.dimensions.ai/details/publication/pub.1030895778"
    ], 
    "sdDataset": "chapters", 
    "sdDatePublished": "2022-11-24T21:20", 
    "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/chapter/chapter_82.jsonl", 
    "type": "Chapter", 
    "url": "https://doi.org/10.1007/978-1-4471-4640-7_8"
  }
]
 

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.10.1007/978-1-4471-4640-7_8'

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

curl -H 'Accept: application/n-triples' 'https://scigraph.springernature.com/pub.10.1007/978-1-4471-4640-7_8'

Turtle is a human-readable linked data format.

curl -H 'Accept: text/turtle' 'https://scigraph.springernature.com/pub.10.1007/978-1-4471-4640-7_8'

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

curl -H 'Accept: application/rdf+xml' 'https://scigraph.springernature.com/pub.10.1007/978-1-4471-4640-7_8'


 

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

185 TRIPLES      22 PREDICATES      86 URIs      79 LITERALS      7 BLANK NODES

Subject Predicate Object
1 sg:pub.10.1007/978-1-4471-4640-7_8 schema:about anzsrc-for:08
2 anzsrc-for:0801
3 schema:author N6e837171c4604678aa2cf447fbcf24c4
4 schema:datePublished 2013
5 schema:datePublishedReg 2013-01-01
6 schema:description The recent proliferation of the Microsoft Kinect, a cheap but quality depth sensor, has brought the need for a challenging category-level 3D object detection dataset to the forefront. Such a dataset can be used for object recognition in a spirit usually reserved for the large collections of intensity images typically collected from the Internet. Here, we will review current 3D datasets and find them lacking in variation of scene, category, instance, and viewpoint. The Berkeley 3D Object Dataset (B3DO), which contains color and depth image pairs gathered in read domestic and office environments will be presented. Baseline object recognition performance in a PASCAL VOC-style detection task is established, and two ways that inferred world size of the object van be used to improve detection are suggested. In an effort to make more significant performance progress, the problem of extracting useful features from range images is addressed. There has been much success in using the histogram of oriented gradients (HOG) as a global descriptor for object detection in intensity images. There are also many proposed descriptors designed specifically for depth data (spin images, shape context, etc.), but these are often focused on the local, not global descriptor paradigm. We explore the failures of gradient-based descriptors when applied to depth, and propose that the proper global descriptor in the realm of 3D should be based on curvature, not gradients.
7 schema:editor N60956208c1c44670ae22a5d8f570b9de
8 schema:genre chapter
9 schema:isAccessibleForFree false
10 schema:isPartOf N588399f8ffd24c61829fd0ab9c85ee42
11 schema:keywords Internet
12 Kinect
13 Microsoft Kinect
14 Van
15 categories
16 collection
17 color
18 curvature
19 data
20 dataset
21 depth
22 depth data
23 depth image pairs
24 depth sensor
25 descriptors
26 detection
27 detection task
28 efforts
29 environment
30 failure
31 features
32 forefront
33 global descriptors
34 gradient
35 gradient-based descriptors
36 histogram
37 image pairs
38 images
39 instances
40 intensity images
41 large collection
42 need
43 object dataset
44 object detection
45 object recognition
46 object recognition performance
47 office environment
48 pairs
49 paradigm
50 performance
51 performance progress
52 problem
53 progress
54 proliferation
55 range images
56 realm
57 recent proliferation
58 recognition
59 recognition performance
60 scene
61 sensors
62 size
63 spirit
64 success
65 task
66 useful features
67 variation
68 variations of scenes
69 viewpoint
70 way
71 world size
72 schema:name A Category-Level 3D Object Dataset: Putting the Kinect to Work
73 schema:pagination 141-165
74 schema:productId N9e369d30348e4be9909f9fefef406004
75 Nb2d5313f27ce40bca676a33e3590e324
76 schema:publisher Nbd786aff7d7043b8b6e8289fce4b61fa
77 schema:sameAs https://app.dimensions.ai/details/publication/pub.1030895778
78 https://doi.org/10.1007/978-1-4471-4640-7_8
79 schema:sdDatePublished 2022-11-24T21:20
80 schema:sdLicense https://scigraph.springernature.com/explorer/license/
81 schema:sdPublisher N1598193042d7434d9918551aa2b4368e
82 schema:url https://doi.org/10.1007/978-1-4471-4640-7_8
83 sgo:license sg:explorer/license/
84 sgo:sdDataset chapters
85 rdf:type schema:Chapter
86 N1598193042d7434d9918551aa2b4368e schema:name Springer Nature - SN SciGraph project
87 rdf:type schema:Organization
88 N1bae558710394179871edb0a7d13d03b schema:familyName Fossati
89 schema:givenName Andrea
90 rdf:type schema:Person
91 N1cd5c92fa98f4843bbda6474d99d50f2 rdf:first sg:person.013361072755.17
92 rdf:rest Na1eb62d482e84080835a8aec6625ef6c
93 N4c3f067084a549bf9b8d82a37634412a schema:familyName Grabner
94 schema:givenName Helmut
95 rdf:type schema:Person
96 N588399f8ffd24c61829fd0ab9c85ee42 schema:isbn 978-1-4471-4639-1
97 978-1-4471-4640-7
98 schema:name Consumer Depth Cameras for Computer Vision
99 rdf:type schema:Book
100 N60956208c1c44670ae22a5d8f570b9de rdf:first N1bae558710394179871edb0a7d13d03b
101 rdf:rest Nee32fc9646dc45ea82fe0513e720cf2a
102 N6e837171c4604678aa2cf447fbcf24c4 rdf:first sg:person.014576140257.12
103 rdf:rest Nd742b5ea047c4c6288ecd27d84393c0c
104 N71123831cd574476b7d5ae510904df5a schema:familyName Konolige
105 schema:givenName Kurt
106 rdf:type schema:Person
107 N7b49715e3b9b450cad6e35fc74085f76 rdf:first N4c3f067084a549bf9b8d82a37634412a
108 rdf:rest Ncf6e4b0e1ddc40c68502a9a5c2e323c6
109 N84c20339591045a1a9d03c53f22f5bce rdf:first N71123831cd574476b7d5ae510904df5a
110 rdf:rest rdf:nil
111 N9e369d30348e4be9909f9fefef406004 schema:name doi
112 schema:value 10.1007/978-1-4471-4640-7_8
113 rdf:type schema:PropertyValue
114 Na1eb62d482e84080835a8aec6625ef6c rdf:first sg:person.0617252060.44
115 rdf:rest Nfe0801cc10464851b998249767720328
116 Nb2d5313f27ce40bca676a33e3590e324 schema:name dimensions_id
117 schema:value pub.1030895778
118 rdf:type schema:PropertyValue
119 Nb629ae88d85d4b4c976b094613e49c74 schema:familyName Gall
120 schema:givenName Juergen
121 rdf:type schema:Person
122 Nbd786aff7d7043b8b6e8289fce4b61fa schema:name Springer Nature
123 rdf:type schema:Organisation
124 Ncf6e4b0e1ddc40c68502a9a5c2e323c6 rdf:first Ne620a29023c349bd834b7a4bb4f7e90e
125 rdf:rest N84c20339591045a1a9d03c53f22f5bce
126 Nd742b5ea047c4c6288ecd27d84393c0c rdf:first sg:person.016171101257.73
127 rdf:rest Ne9b8dea87659419eb153706bea13ea0c
128 Ne3e58533c85e4530b080d3c68610181c rdf:first sg:person.01302212713.98
129 rdf:rest N1cd5c92fa98f4843bbda6474d99d50f2
130 Ne620a29023c349bd834b7a4bb4f7e90e schema:familyName Ren
131 schema:givenName Xiaofeng
132 rdf:type schema:Person
133 Ne9b8dea87659419eb153706bea13ea0c rdf:first sg:person.013220623467.72
134 rdf:rest Ne3e58533c85e4530b080d3c68610181c
135 Nee32fc9646dc45ea82fe0513e720cf2a rdf:first Nb629ae88d85d4b4c976b094613e49c74
136 rdf:rest N7b49715e3b9b450cad6e35fc74085f76
137 Nfe0801cc10464851b998249767720328 rdf:first sg:person.01001613660.25
138 rdf:rest rdf:nil
139 anzsrc-for:08 schema:inDefinedTermSet anzsrc-for:
140 schema:name Information and Computing Sciences
141 rdf:type schema:DefinedTerm
142 anzsrc-for:0801 schema:inDefinedTermSet anzsrc-for:
143 schema:name Artificial Intelligence and Image Processing
144 rdf:type schema:DefinedTerm
145 sg:person.01001613660.25 schema:affiliation grid-institutes:grid.47840.3f
146 schema:familyName Darrell
147 schema:givenName Trevor
148 schema:sameAs https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.01001613660.25
149 rdf:type schema:Person
150 sg:person.01302212713.98 schema:affiliation grid-institutes:grid.47840.3f
151 schema:familyName Barron
152 schema:givenName Jonathan T.
153 schema:sameAs https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.01302212713.98
154 rdf:type schema:Person
155 sg:person.013220623467.72 schema:affiliation grid-institutes:grid.47840.3f
156 schema:familyName Jia
157 schema:givenName Yangqing
158 schema:sameAs https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.013220623467.72
159 rdf:type schema:Person
160 sg:person.013361072755.17 schema:affiliation grid-institutes:None
161 schema:familyName Fritz
162 schema:givenName Mario
163 schema:sameAs https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.013361072755.17
164 rdf:type schema:Person
165 sg:person.014576140257.12 schema:affiliation grid-institutes:grid.47840.3f
166 schema:familyName Janoch
167 schema:givenName Allison
168 schema:sameAs https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.014576140257.12
169 rdf:type schema:Person
170 sg:person.016171101257.73 schema:affiliation grid-institutes:grid.47840.3f
171 schema:familyName Karayev
172 schema:givenName Sergey
173 schema:sameAs https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.016171101257.73
174 rdf:type schema:Person
175 sg:person.0617252060.44 schema:affiliation grid-institutes:grid.47840.3f
176 schema:familyName Saenko
177 schema:givenName Kate
178 schema:sameAs https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.0617252060.44
179 rdf:type schema:Person
180 grid-institutes:None schema:alternateName Max Plank Institute for Informatics, Campus E1.4, 66123, Saarbrücken, Germany
181 schema:name Max Plank Institute for Informatics, Campus E1.4, 66123, Saarbrücken, Germany
182 rdf:type schema:Organization
183 grid-institutes:grid.47840.3f schema:alternateName University of California at Berkeley, Berkeley, CA, USA
184 schema:name University of California at Berkeley, Berkeley, CA, USA
185 rdf:type schema:Organization
 




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


...