Interpolating Convolutional Neural Networks Using Batch Normalization View Full Text


Ontology type: schema:Chapter     


Chapter Info

DATE

2018-10-06

AUTHORS

Gratianus Wesley Putra Data , Kirjon Ngu , David William Murray , Victor Adrian Prisacariu

ABSTRACT

Perceiving a visual concept as a mixture of learned ones is natural for humans, aiding them to grasp new concepts and strengthening old ones. For all their power and recent success, deep convolutional networks do not have this ability. Inspired by recent work on universal representations for neural networks, we propose a simple emulation of this mechanism by purposing batch normalization layers to discriminate visual classes, and formulating a way to combine them to solve new tasks. We show that this can be applied for 2-way few-shot learning where we obtain between 4% and 17% better accuracy compared to straightforward full fine-tuning, and demonstrate that it can also be extended to the orthogonal application of style transfer. More... »

PAGES

591-606

Book

TITLE

Computer Vision – ECCV 2018

ISBN

978-3-030-01260-1
978-3-030-01261-8

Author Affiliations

Identifiers

URI

http://scigraph.springernature.com/pub.10.1007/978-3-030-01261-8_35

DOI

http://dx.doi.org/10.1007/978-3-030-01261-8_35

DIMENSIONS

https://app.dimensions.ai/details/publication/pub.1107502687


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/2008/0801", 
        "inDefinedTermSet": "http://purl.org/au-research/vocabulary/anzsrc-for/2008/", 
        "name": "Artificial Intelligence and Image Processing", 
        "type": "DefinedTerm"
      }, 
      {
        "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"
      }
    ], 
    "author": [
      {
        "affiliation": {
          "alternateName": "University of Oxford", 
          "id": "https://www.grid.ac/institutes/grid.4991.5", 
          "name": [
            "Active Vision Laboratory, Department of Engineering Science, University of Oxford, Oxford, UK"
          ], 
          "type": "Organization"
        }, 
        "familyName": "Data", 
        "givenName": "Gratianus Wesley Putra", 
        "id": "sg:person.015105667325.16", 
        "sameAs": [
          "https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.015105667325.16"
        ], 
        "type": "Person"
      }, 
      {
        "affiliation": {
          "alternateName": "University of Oxford", 
          "id": "https://www.grid.ac/institutes/grid.4991.5", 
          "name": [
            "Active Vision Laboratory, Department of Engineering Science, University of Oxford, Oxford, UK"
          ], 
          "type": "Organization"
        }, 
        "familyName": "Ngu", 
        "givenName": "Kirjon", 
        "type": "Person"
      }, 
      {
        "affiliation": {
          "alternateName": "University of Oxford", 
          "id": "https://www.grid.ac/institutes/grid.4991.5", 
          "name": [
            "Active Vision Laboratory, Department of Engineering Science, University of Oxford, Oxford, UK"
          ], 
          "type": "Organization"
        }, 
        "familyName": "Murray", 
        "givenName": "David William", 
        "id": "sg:person.012633757262.69", 
        "sameAs": [
          "https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.012633757262.69"
        ], 
        "type": "Person"
      }, 
      {
        "affiliation": {
          "alternateName": "University of Oxford", 
          "id": "https://www.grid.ac/institutes/grid.4991.5", 
          "name": [
            "Active Vision Laboratory, Department of Engineering Science, University of Oxford, Oxford, UK"
          ], 
          "type": "Organization"
        }, 
        "familyName": "Prisacariu", 
        "givenName": "Victor Adrian", 
        "id": "sg:person.011045366441.02", 
        "sameAs": [
          "https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.011045366441.02"
        ], 
        "type": "Person"
      }
    ], 
    "citation": [
      {
        "id": "https://doi.org/10.1109/jproc.2010.2040551", 
        "sameAs": [
          "https://app.dimensions.ai/details/publication/pub.1061297220"
        ], 
        "type": "CreativeWork"
      }, 
      {
        "id": "https://doi.org/10.1109/tsp.2006.881199", 
        "sameAs": [
          "https://app.dimensions.ai/details/publication/pub.1061800223"
        ], 
        "type": "CreativeWork"
      }, 
      {
        "id": "https://doi.org/10.1145/3065386", 
        "sameAs": [
          "https://app.dimensions.ai/details/publication/pub.1085642448"
        ], 
        "type": "CreativeWork"
      }, 
      {
        "id": "https://doi.org/10.1145/3065386", 
        "sameAs": [
          "https://app.dimensions.ai/details/publication/pub.1085642448"
        ], 
        "type": "CreativeWork"
      }, 
      {
        "id": "https://doi.org/10.1109/cvpr.2016.90", 
        "sameAs": [
          "https://app.dimensions.ai/details/publication/pub.1093359587"
        ], 
        "type": "CreativeWork"
      }, 
      {
        "id": "https://doi.org/10.1109/icassp.1999.760624", 
        "sameAs": [
          "https://app.dimensions.ai/details/publication/pub.1093656198"
        ], 
        "type": "CreativeWork"
      }, 
      {
        "id": "https://doi.org/10.1109/iccv.2017.328", 
        "sameAs": [
          "https://app.dimensions.ai/details/publication/pub.1100060313"
        ], 
        "type": "CreativeWork"
      }
    ], 
    "datePublished": "2018-10-06", 
    "datePublishedReg": "2018-10-06", 
    "description": "Perceiving a visual concept as a mixture of learned ones is natural for humans, aiding them to grasp new concepts and strengthening old ones. For all their power and recent success, deep convolutional networks do not have this ability. Inspired by recent work on universal representations for neural networks, we propose a simple emulation of this mechanism by purposing batch normalization layers to discriminate visual classes, and formulating a way to combine them to solve new tasks. We show that this can be applied for 2-way few-shot learning where we obtain between 4% and 17% better accuracy compared to straightforward full fine-tuning, and demonstrate that it can also be extended to the orthogonal application of style transfer.", 
    "editor": [
      {
        "familyName": "Ferrari", 
        "givenName": "Vittorio", 
        "type": "Person"
      }, 
      {
        "familyName": "Hebert", 
        "givenName": "Martial", 
        "type": "Person"
      }, 
      {
        "familyName": "Sminchisescu", 
        "givenName": "Cristian", 
        "type": "Person"
      }, 
      {
        "familyName": "Weiss", 
        "givenName": "Yair", 
        "type": "Person"
      }
    ], 
    "genre": "chapter", 
    "id": "sg:pub.10.1007/978-3-030-01261-8_35", 
    "inLanguage": [
      "en"
    ], 
    "isAccessibleForFree": false, 
    "isPartOf": {
      "isbn": [
        "978-3-030-01260-1", 
        "978-3-030-01261-8"
      ], 
      "name": "Computer Vision \u2013 ECCV 2018", 
      "type": "Book"
    }, 
    "name": "Interpolating Convolutional Neural Networks Using Batch Normalization", 
    "pagination": "591-606", 
    "productId": [
      {
        "name": "doi", 
        "type": "PropertyValue", 
        "value": [
          "10.1007/978-3-030-01261-8_35"
        ]
      }, 
      {
        "name": "readcube_id", 
        "type": "PropertyValue", 
        "value": [
          "e14fb0691474b5fcabbdc09c1f1b1a7a6f3c3da01f7ae305d90bd1b76e3d8016"
        ]
      }, 
      {
        "name": "dimensions_id", 
        "type": "PropertyValue", 
        "value": [
          "pub.1107502687"
        ]
      }
    ], 
    "publisher": {
      "location": "Cham", 
      "name": "Springer International Publishing", 
      "type": "Organisation"
    }, 
    "sameAs": [
      "https://doi.org/10.1007/978-3-030-01261-8_35", 
      "https://app.dimensions.ai/details/publication/pub.1107502687"
    ], 
    "sdDataset": "chapters", 
    "sdDatePublished": "2019-04-16T04:39", 
    "sdLicense": "https://scigraph.springernature.com/explorer/license/", 
    "sdPublisher": {
      "name": "Springer Nature - SN SciGraph project", 
      "type": "Organization"
    }, 
    "sdSource": "s3://com-uberresearch-data-dimensions-target-20181106-alternative/cleanup/v134/2549eaecd7973599484d7c17b260dba0a4ecb94b/merge/v9/a6c9fde33151104705d4d7ff012ea9563521a3ce/jats-lookup/v90/0000000321_0000000321/records_74903_00000000.jsonl", 
    "type": "Chapter", 
    "url": "https://link.springer.com/10.1007%2F978-3-030-01261-8_35"
  }
]
 

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.10.1007/978-3-030-01261-8_35'

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

curl -H 'Accept: application/n-triples' 'https://scigraph.springernature.com/pub.10.1007/978-3-030-01261-8_35'

Turtle is a human-readable linked data format.

curl -H 'Accept: text/turtle' 'https://scigraph.springernature.com/pub.10.1007/978-3-030-01261-8_35'

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

curl -H 'Accept: application/rdf+xml' 'https://scigraph.springernature.com/pub.10.1007/978-3-030-01261-8_35'


 

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

118 TRIPLES      23 PREDICATES      32 URIs      19 LITERALS      8 BLANK NODES

Subject Predicate Object
1 sg:pub.10.1007/978-3-030-01261-8_35 schema:about anzsrc-for:08
2 anzsrc-for:0801
3 schema:author N950856c245de40a7a6661785e056b33c
4 schema:citation https://doi.org/10.1109/cvpr.2016.90
5 https://doi.org/10.1109/icassp.1999.760624
6 https://doi.org/10.1109/iccv.2017.328
7 https://doi.org/10.1109/jproc.2010.2040551
8 https://doi.org/10.1109/tsp.2006.881199
9 https://doi.org/10.1145/3065386
10 schema:datePublished 2018-10-06
11 schema:datePublishedReg 2018-10-06
12 schema:description Perceiving a visual concept as a mixture of learned ones is natural for humans, aiding them to grasp new concepts and strengthening old ones. For all their power and recent success, deep convolutional networks do not have this ability. Inspired by recent work on universal representations for neural networks, we propose a simple emulation of this mechanism by purposing batch normalization layers to discriminate visual classes, and formulating a way to combine them to solve new tasks. We show that this can be applied for 2-way few-shot learning where we obtain between 4% and 17% better accuracy compared to straightforward full fine-tuning, and demonstrate that it can also be extended to the orthogonal application of style transfer.
13 schema:editor N0714e11194704f60bc5f4cfa9fb15fa8
14 schema:genre chapter
15 schema:inLanguage en
16 schema:isAccessibleForFree false
17 schema:isPartOf N04217ff697074f8188094f1383d572d4
18 schema:name Interpolating Convolutional Neural Networks Using Batch Normalization
19 schema:pagination 591-606
20 schema:productId N2d19d9d0b5294fcfb625c2d04b30533c
21 N3835203ad8054460a909807e58547919
22 N8726f15ef0994aa8967801205468032f
23 schema:publisher Nf7f3ea2f42324c3bbf8d45394122db73
24 schema:sameAs https://app.dimensions.ai/details/publication/pub.1107502687
25 https://doi.org/10.1007/978-3-030-01261-8_35
26 schema:sdDatePublished 2019-04-16T04:39
27 schema:sdLicense https://scigraph.springernature.com/explorer/license/
28 schema:sdPublisher N250ab02dee104fda82032a475fdd172c
29 schema:url https://link.springer.com/10.1007%2F978-3-030-01261-8_35
30 sgo:license sg:explorer/license/
31 sgo:sdDataset chapters
32 rdf:type schema:Chapter
33 N04217ff697074f8188094f1383d572d4 schema:isbn 978-3-030-01260-1
34 978-3-030-01261-8
35 schema:name Computer Vision – ECCV 2018
36 rdf:type schema:Book
37 N0714e11194704f60bc5f4cfa9fb15fa8 rdf:first Nab7cc49d673c4931b8be1663e242f45f
38 rdf:rest Nb5f8fb92ca5e4281acd9ced89b3975f2
39 N0c619a7fea094c90b9bf7592a6509e6c rdf:first N74c3eb3b37ea41e0a5cad283fb040041
40 rdf:rest N66ddd54c08cb478a8c0b917c821a4e5a
41 N250ab02dee104fda82032a475fdd172c schema:name Springer Nature - SN SciGraph project
42 rdf:type schema:Organization
43 N2d19d9d0b5294fcfb625c2d04b30533c schema:name doi
44 schema:value 10.1007/978-3-030-01261-8_35
45 rdf:type schema:PropertyValue
46 N3835203ad8054460a909807e58547919 schema:name readcube_id
47 schema:value e14fb0691474b5fcabbdc09c1f1b1a7a6f3c3da01f7ae305d90bd1b76e3d8016
48 rdf:type schema:PropertyValue
49 N4f7b988f6a49446c8e1c1720972f4745 rdf:first sg:person.011045366441.02
50 rdf:rest rdf:nil
51 N62393ad08c874a43810e5aa33a43e13c rdf:first N65f5c557affe450ab99966baa7afe4de
52 rdf:rest rdf:nil
53 N65f5c557affe450ab99966baa7afe4de schema:familyName Weiss
54 schema:givenName Yair
55 rdf:type schema:Person
56 N66ddd54c08cb478a8c0b917c821a4e5a rdf:first sg:person.012633757262.69
57 rdf:rest N4f7b988f6a49446c8e1c1720972f4745
58 N74c3eb3b37ea41e0a5cad283fb040041 schema:affiliation https://www.grid.ac/institutes/grid.4991.5
59 schema:familyName Ngu
60 schema:givenName Kirjon
61 rdf:type schema:Person
62 N8726f15ef0994aa8967801205468032f schema:name dimensions_id
63 schema:value pub.1107502687
64 rdf:type schema:PropertyValue
65 N8a4a10a46ae144fa9365e427ac87858a schema:familyName Sminchisescu
66 schema:givenName Cristian
67 rdf:type schema:Person
68 N950856c245de40a7a6661785e056b33c rdf:first sg:person.015105667325.16
69 rdf:rest N0c619a7fea094c90b9bf7592a6509e6c
70 Nab7cc49d673c4931b8be1663e242f45f schema:familyName Ferrari
71 schema:givenName Vittorio
72 rdf:type schema:Person
73 Nb478eaad62bd418f911660d4a28fa8f0 rdf:first N8a4a10a46ae144fa9365e427ac87858a
74 rdf:rest N62393ad08c874a43810e5aa33a43e13c
75 Nb5f8fb92ca5e4281acd9ced89b3975f2 rdf:first Ndde310776dc5457a9d76c5a21e652696
76 rdf:rest Nb478eaad62bd418f911660d4a28fa8f0
77 Ndde310776dc5457a9d76c5a21e652696 schema:familyName Hebert
78 schema:givenName Martial
79 rdf:type schema:Person
80 Nf7f3ea2f42324c3bbf8d45394122db73 schema:location Cham
81 schema:name Springer International Publishing
82 rdf:type schema:Organisation
83 anzsrc-for:08 schema:inDefinedTermSet anzsrc-for:
84 schema:name Information and Computing Sciences
85 rdf:type schema:DefinedTerm
86 anzsrc-for:0801 schema:inDefinedTermSet anzsrc-for:
87 schema:name Artificial Intelligence and Image Processing
88 rdf:type schema:DefinedTerm
89 sg:person.011045366441.02 schema:affiliation https://www.grid.ac/institutes/grid.4991.5
90 schema:familyName Prisacariu
91 schema:givenName Victor Adrian
92 schema:sameAs https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.011045366441.02
93 rdf:type schema:Person
94 sg:person.012633757262.69 schema:affiliation https://www.grid.ac/institutes/grid.4991.5
95 schema:familyName Murray
96 schema:givenName David William
97 schema:sameAs https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.012633757262.69
98 rdf:type schema:Person
99 sg:person.015105667325.16 schema:affiliation https://www.grid.ac/institutes/grid.4991.5
100 schema:familyName Data
101 schema:givenName Gratianus Wesley Putra
102 schema:sameAs https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.015105667325.16
103 rdf:type schema:Person
104 https://doi.org/10.1109/cvpr.2016.90 schema:sameAs https://app.dimensions.ai/details/publication/pub.1093359587
105 rdf:type schema:CreativeWork
106 https://doi.org/10.1109/icassp.1999.760624 schema:sameAs https://app.dimensions.ai/details/publication/pub.1093656198
107 rdf:type schema:CreativeWork
108 https://doi.org/10.1109/iccv.2017.328 schema:sameAs https://app.dimensions.ai/details/publication/pub.1100060313
109 rdf:type schema:CreativeWork
110 https://doi.org/10.1109/jproc.2010.2040551 schema:sameAs https://app.dimensions.ai/details/publication/pub.1061297220
111 rdf:type schema:CreativeWork
112 https://doi.org/10.1109/tsp.2006.881199 schema:sameAs https://app.dimensions.ai/details/publication/pub.1061800223
113 rdf:type schema:CreativeWork
114 https://doi.org/10.1145/3065386 schema:sameAs https://app.dimensions.ai/details/publication/pub.1085642448
115 rdf:type schema:CreativeWork
116 https://www.grid.ac/institutes/grid.4991.5 schema:alternateName University of Oxford
117 schema:name Active Vision Laboratory, Department of Engineering Science, University of Oxford, Oxford, UK
118 rdf:type schema:Organization
 




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


...