Partial Adversarial Domain Adaptation View Full Text


Ontology type: schema:Chapter      Open Access: True


Chapter Info

DATE

2018-10-07

AUTHORS

Zhangjie Cao , Lijia Ma , Mingsheng Long , Jianmin Wang

ABSTRACT

Domain adversarial learning aligns the feature distributions across the source and target domains in a two-player minimax game. Existing domain adversarial networks generally assume identical label space across different domains. In the presence of big data, there is strong motivation of transferring deep models from existing big domains to unknown small domains. This paper introduces partial domain adaptation as a new domain adaptation scenario, which relaxes the fully shared label space assumption to that the source label space subsumes the target label space. Previous methods typically match the whole source domain to the target domain, which are vulnerable to negative transfer for the partial domain adaptation problem due to the large mismatch between label spaces. We present Partial Adversarial Domain Adaptation (PADA), which simultaneously alleviates negative transfer by down-weighing the data of outlier source classes for training both source classifier and domain adversary, and promotes positive transfer by matching the feature distributions in the shared label space. Experiments show that PADA exceeds state-of-the-art results for partial domain adaptation tasks on several datasets. More... »

PAGES

139-155

Identifiers

URI

http://scigraph.springernature.com/pub.10.1007/978-3-030-01237-3_9

DOI

http://dx.doi.org/10.1007/978-3-030-01237-3_9

DIMENSIONS

https://app.dimensions.ai/details/publication/pub.1107463340


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": "National Engineering Laboratory for Big Data Software, Beijing National Research Center for Information Science and Technology, Beijing, China", 
          "id": "http://www.grid.ac/institutes/None", 
          "name": [
            "School of Software, Tsinghua University, Beijing, China", 
            "National Engineering Laboratory for Big Data Software, Beijing National Research Center for Information Science and Technology, Beijing, China"
          ], 
          "type": "Organization"
        }, 
        "familyName": "Cao", 
        "givenName": "Zhangjie", 
        "id": "sg:person.013022415410.38", 
        "sameAs": [
          "https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.013022415410.38"
        ], 
        "type": "Person"
      }, 
      {
        "affiliation": {
          "alternateName": "National Engineering Laboratory for Big Data Software, Beijing National Research Center for Information Science and Technology, Beijing, China", 
          "id": "http://www.grid.ac/institutes/None", 
          "name": [
            "School of Software, Tsinghua University, Beijing, China", 
            "National Engineering Laboratory for Big Data Software, Beijing National Research Center for Information Science and Technology, Beijing, China"
          ], 
          "type": "Organization"
        }, 
        "familyName": "Ma", 
        "givenName": "Lijia", 
        "type": "Person"
      }, 
      {
        "affiliation": {
          "alternateName": "National Engineering Laboratory for Big Data Software, Beijing National Research Center for Information Science and Technology, Beijing, China", 
          "id": "http://www.grid.ac/institutes/None", 
          "name": [
            "School of Software, Tsinghua University, Beijing, China", 
            "National Engineering Laboratory for Big Data Software, Beijing National Research Center for Information Science and Technology, Beijing, China"
          ], 
          "type": "Organization"
        }, 
        "familyName": "Long", 
        "givenName": "Mingsheng", 
        "id": "sg:person.013417115303.81", 
        "sameAs": [
          "https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.013417115303.81"
        ], 
        "type": "Person"
      }, 
      {
        "affiliation": {
          "alternateName": "National Engineering Laboratory for Big Data Software, Beijing National Research Center for Information Science and Technology, Beijing, China", 
          "id": "http://www.grid.ac/institutes/None", 
          "name": [
            "School of Software, Tsinghua University, Beijing, China", 
            "National Engineering Laboratory for Big Data Software, Beijing National Research Center for Information Science and Technology, Beijing, China"
          ], 
          "type": "Organization"
        }, 
        "familyName": "Wang", 
        "givenName": "Jianmin", 
        "id": "sg:person.012303351315.43", 
        "sameAs": [
          "https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.012303351315.43"
        ], 
        "type": "Person"
      }
    ], 
    "datePublished": "2018-10-07", 
    "datePublishedReg": "2018-10-07", 
    "description": "Domain adversarial learning aligns the feature distributions across the source and target domains in a two-player minimax game. Existing domain adversarial networks generally assume identical label space across different domains. In the presence of big data, there is strong motivation of transferring deep models from existing big domains to unknown small domains. This paper introduces partial domain adaptation as a new domain adaptation scenario, which relaxes the fully shared label space assumption to that the source label space subsumes the target label space. Previous methods typically match the whole source domain\u00a0to the target domain, which are vulnerable to negative transfer for the partial domain adaptation problem due to the large mismatch between label spaces. We present Partial Adversarial Domain Adaptation (PADA), which simultaneously alleviates negative transfer by down-weighing the data of outlier source classes for training both source classifier and domain adversary, and promotes positive transfer by matching the feature distributions in the shared label space. Experiments show that PADA exceeds state-of-the-art results for partial domain adaptation tasks on several datasets.", 
    "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.10.1007/978-3-030-01237-3_9", 
    "inLanguage": "en", 
    "isAccessibleForFree": true, 
    "isPartOf": {
      "isbn": [
        "978-3-030-01236-6", 
        "978-3-030-01237-3"
      ], 
      "name": "Computer Vision \u2013 ECCV 2018", 
      "type": "Book"
    }, 
    "keywords": [
      "Adversarial Domain Adaptation", 
      "label space", 
      "domain adaptation", 
      "target domain", 
      "feature distributions", 
      "domain adaptation problem", 
      "domain adaptation scenarios", 
      "domain adaptation tasks", 
      "adversarial learning", 
      "Adversarial Networks", 
      "deep model", 
      "big data", 
      "art results", 
      "source domain", 
      "source classifier", 
      "adaptation problem", 
      "minimax game", 
      "adaptation scenarios", 
      "whole source domain", 
      "different domains", 
      "previous methods", 
      "negative transfer", 
      "adaptation task", 
      "source classes", 
      "adversary", 
      "Big domain", 
      "classifier", 
      "domain", 
      "space", 
      "datasets", 
      "network", 
      "task", 
      "learning", 
      "scenarios", 
      "game", 
      "strong motivation", 
      "adaptation", 
      "data", 
      "training", 
      "model", 
      "method", 
      "class", 
      "experiments", 
      "small domains", 
      "motivation", 
      "mismatch", 
      "assumption", 
      "results", 
      "state", 
      "source", 
      "distribution", 
      "transfer", 
      "positive transfer", 
      "large mismatch", 
      "presence", 
      "paper", 
      "problem", 
      "two-player minimax game", 
      "domain adversarial networks", 
      "identical label space", 
      "unknown small domains", 
      "partial domain adaptation", 
      "new domain adaptation scenario", 
      "label space assumption", 
      "space assumption", 
      "source label space", 
      "target label space", 
      "partial domain adaptation problem", 
      "Partial Adversarial Domain Adaptation", 
      "outlier source classes", 
      "domain adversary", 
      "partial domain adaptation tasks"
    ], 
    "name": "Partial Adversarial Domain Adaptation", 
    "pagination": "139-155", 
    "productId": [
      {
        "name": "dimensions_id", 
        "type": "PropertyValue", 
        "value": [
          "pub.1107463340"
        ]
      }, 
      {
        "name": "doi", 
        "type": "PropertyValue", 
        "value": [
          "10.1007/978-3-030-01237-3_9"
        ]
      }
    ], 
    "publisher": {
      "name": "Springer Nature", 
      "type": "Organisation"
    }, 
    "sameAs": [
      "https://doi.org/10.1007/978-3-030-01237-3_9", 
      "https://app.dimensions.ai/details/publication/pub.1107463340"
    ], 
    "sdDataset": "chapters", 
    "sdDatePublished": "2022-01-01T19:11", 
    "sdLicense": "https://scigraph.springernature.com/explorer/license/", 
    "sdPublisher": {
      "name": "Springer Nature - SN SciGraph project", 
      "type": "Organization"
    }, 
    "sdSource": "s3://com-springernature-scigraph/baseset/20220101/entities/gbq_results/chapter/chapter_189.jsonl", 
    "type": "Chapter", 
    "url": "https://doi.org/10.1007/978-3-030-01237-3_9"
  }
]
 

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.10.1007/978-3-030-01237-3_9'

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

curl -H 'Accept: application/n-triples' 'https://scigraph.springernature.com/pub.10.1007/978-3-030-01237-3_9'

Turtle is a human-readable linked data format.

curl -H 'Accept: text/turtle' 'https://scigraph.springernature.com/pub.10.1007/978-3-030-01237-3_9'

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

curl -H 'Accept: application/rdf+xml' 'https://scigraph.springernature.com/pub.10.1007/978-3-030-01237-3_9'


 

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

168 TRIPLES      23 PREDICATES      96 URIs      89 LITERALS      7 BLANK NODES

Subject Predicate Object
1 sg:pub.10.1007/978-3-030-01237-3_9 schema:about anzsrc-for:08
2 anzsrc-for:0801
3 schema:author N4704b161950540b88ebbb684ffa0c676
4 schema:datePublished 2018-10-07
5 schema:datePublishedReg 2018-10-07
6 schema:description Domain adversarial learning aligns the feature distributions across the source and target domains in a two-player minimax game. Existing domain adversarial networks generally assume identical label space across different domains. In the presence of big data, there is strong motivation of transferring deep models from existing big domains to unknown small domains. This paper introduces partial domain adaptation as a new domain adaptation scenario, which relaxes the fully shared label space assumption to that the source label space subsumes the target label space. Previous methods typically match the whole source domain to the target domain, which are vulnerable to negative transfer for the partial domain adaptation problem due to the large mismatch between label spaces. We present Partial Adversarial Domain Adaptation (PADA), which simultaneously alleviates negative transfer by down-weighing the data of outlier source classes for training both source classifier and domain adversary, and promotes positive transfer by matching the feature distributions in the shared label space. Experiments show that PADA exceeds state-of-the-art results for partial domain adaptation tasks on several datasets.
7 schema:editor Nf3d1e1644b2649a791e303181c5c2197
8 schema:genre chapter
9 schema:inLanguage en
10 schema:isAccessibleForFree true
11 schema:isPartOf Nc52fdf9842ac4014b5e162a674838668
12 schema:keywords Adversarial Domain Adaptation
13 Adversarial Networks
14 Big domain
15 Partial Adversarial Domain Adaptation
16 adaptation
17 adaptation problem
18 adaptation scenarios
19 adaptation task
20 adversarial learning
21 adversary
22 art results
23 assumption
24 big data
25 class
26 classifier
27 data
28 datasets
29 deep model
30 different domains
31 distribution
32 domain
33 domain adaptation
34 domain adaptation problem
35 domain adaptation scenarios
36 domain adaptation tasks
37 domain adversarial networks
38 domain adversary
39 experiments
40 feature distributions
41 game
42 identical label space
43 label space
44 label space assumption
45 large mismatch
46 learning
47 method
48 minimax game
49 mismatch
50 model
51 motivation
52 negative transfer
53 network
54 new domain adaptation scenario
55 outlier source classes
56 paper
57 partial domain adaptation
58 partial domain adaptation problem
59 partial domain adaptation tasks
60 positive transfer
61 presence
62 previous methods
63 problem
64 results
65 scenarios
66 small domains
67 source
68 source classes
69 source classifier
70 source domain
71 source label space
72 space
73 space assumption
74 state
75 strong motivation
76 target domain
77 target label space
78 task
79 training
80 transfer
81 two-player minimax game
82 unknown small domains
83 whole source domain
84 schema:name Partial Adversarial Domain Adaptation
85 schema:pagination 139-155
86 schema:productId N88cb3d9b804f4526afe7e2d156c06d4f
87 Nfc0dda6d68ff4f3a97d8dc369ea2e1bb
88 schema:publisher N9c68c87b49be4b2eb59a7095d8a23e9b
89 schema:sameAs https://app.dimensions.ai/details/publication/pub.1107463340
90 https://doi.org/10.1007/978-3-030-01237-3_9
91 schema:sdDatePublished 2022-01-01T19:11
92 schema:sdLicense https://scigraph.springernature.com/explorer/license/
93 schema:sdPublisher Nbf67aa87025e48bfb5a6658fd9729a53
94 schema:url https://doi.org/10.1007/978-3-030-01237-3_9
95 sgo:license sg:explorer/license/
96 sgo:sdDataset chapters
97 rdf:type schema:Chapter
98 N016f8de41b64481e8518d49a2d10df88 schema:familyName Sminchisescu
99 schema:givenName Cristian
100 rdf:type schema:Person
101 N1902efc3a7364cf5ab89cdbde55faa7e schema:affiliation grid-institutes:None
102 schema:familyName Ma
103 schema:givenName Lijia
104 rdf:type schema:Person
105 N3529477a82b849c3a8b1f2d2671c421d rdf:first sg:person.012303351315.43
106 rdf:rest rdf:nil
107 N3e0de5f7287d4a64b9b61e987a844e5f rdf:first sg:person.013417115303.81
108 rdf:rest N3529477a82b849c3a8b1f2d2671c421d
109 N4704b161950540b88ebbb684ffa0c676 rdf:first sg:person.013022415410.38
110 rdf:rest N874a8c4c7b28457c812ef335dc7cadf8
111 N5b6117f6ccf240548bfadf86aabf359b rdf:first N016f8de41b64481e8518d49a2d10df88
112 rdf:rest N858c20afd9fd4e45b3dcdc8e44b948df
113 N75eb9d4bce024a58abe852bdcec6763f schema:familyName Ferrari
114 schema:givenName Vittorio
115 rdf:type schema:Person
116 N858c20afd9fd4e45b3dcdc8e44b948df rdf:first Nea8c8832daf44556b1a6fa1002bcc728
117 rdf:rest rdf:nil
118 N874a8c4c7b28457c812ef335dc7cadf8 rdf:first N1902efc3a7364cf5ab89cdbde55faa7e
119 rdf:rest N3e0de5f7287d4a64b9b61e987a844e5f
120 N88cb3d9b804f4526afe7e2d156c06d4f schema:name doi
121 schema:value 10.1007/978-3-030-01237-3_9
122 rdf:type schema:PropertyValue
123 N8d29230381ca45f0b9013c46d48bda94 schema:familyName Hebert
124 schema:givenName Martial
125 rdf:type schema:Person
126 N9c68c87b49be4b2eb59a7095d8a23e9b schema:name Springer Nature
127 rdf:type schema:Organisation
128 Nbf67aa87025e48bfb5a6658fd9729a53 schema:name Springer Nature - SN SciGraph project
129 rdf:type schema:Organization
130 Nc3a999baf45649aa890ed5c7a472ba23 rdf:first N8d29230381ca45f0b9013c46d48bda94
131 rdf:rest N5b6117f6ccf240548bfadf86aabf359b
132 Nc52fdf9842ac4014b5e162a674838668 schema:isbn 978-3-030-01236-6
133 978-3-030-01237-3
134 schema:name Computer Vision – ECCV 2018
135 rdf:type schema:Book
136 Nea8c8832daf44556b1a6fa1002bcc728 schema:familyName Weiss
137 schema:givenName Yair
138 rdf:type schema:Person
139 Nf3d1e1644b2649a791e303181c5c2197 rdf:first N75eb9d4bce024a58abe852bdcec6763f
140 rdf:rest Nc3a999baf45649aa890ed5c7a472ba23
141 Nfc0dda6d68ff4f3a97d8dc369ea2e1bb schema:name dimensions_id
142 schema:value pub.1107463340
143 rdf:type schema:PropertyValue
144 anzsrc-for:08 schema:inDefinedTermSet anzsrc-for:
145 schema:name Information and Computing Sciences
146 rdf:type schema:DefinedTerm
147 anzsrc-for:0801 schema:inDefinedTermSet anzsrc-for:
148 schema:name Artificial Intelligence and Image Processing
149 rdf:type schema:DefinedTerm
150 sg:person.012303351315.43 schema:affiliation grid-institutes:None
151 schema:familyName Wang
152 schema:givenName Jianmin
153 schema:sameAs https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.012303351315.43
154 rdf:type schema:Person
155 sg:person.013022415410.38 schema:affiliation grid-institutes:None
156 schema:familyName Cao
157 schema:givenName Zhangjie
158 schema:sameAs https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.013022415410.38
159 rdf:type schema:Person
160 sg:person.013417115303.81 schema:affiliation grid-institutes:None
161 schema:familyName Long
162 schema:givenName Mingsheng
163 schema:sameAs https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.013417115303.81
164 rdf:type schema:Person
165 grid-institutes:None schema:alternateName National Engineering Laboratory for Big Data Software, Beijing National Research Center for Information Science and Technology, Beijing, China
166 schema:name National Engineering Laboratory for Big Data Software, Beijing National Research Center for Information Science and Technology, Beijing, China
167 School of Software, Tsinghua University, Beijing, China
168 rdf:type schema:Organization
 




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


...