Deformable ConvNets V2: More Deformable, Better Results

The superior performance of Deformable Convolutional Networks arises from its ability to adapt to the geometric variations of objects. Through an examination of its adaptive behavior, we observe that while the spatial support for its neural features conforms more closely than regular ConvNets to obj...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:Proceedings (IEEE Computer Society Conference on Computer Vision and Pattern Recognition. Online) s. 9300 - 9308
Hlavní autoři: Zhu, Xizhou, Hu, Han, Lin, Stephen, Dai, Jifeng
Médium: Konferenční příspěvek
Jazyk:angličtina
Vydáno: IEEE 01.06.2019
Témata:
ISSN:1063-6919
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Abstract The superior performance of Deformable Convolutional Networks arises from its ability to adapt to the geometric variations of objects. Through an examination of its adaptive behavior, we observe that while the spatial support for its neural features conforms more closely than regular ConvNets to object structure, this support may nevertheless extend well beyond the region of interest, causing features to be influenced by irrelevant image content. To address this problem, we present a reformulation of Deformable ConvNets that improves its ability to focus on pertinent image regions, through increased modeling power and stronger training. The modeling power is enhanced through a more comprehensive integration of deformable convolution within the network, and by introducing a modulation mechanism that expands the scope of deformation modeling. To effectively harness this enriched modeling capability, we guide network training via a proposed feature mimicking scheme that helps the network to learn features that reflect the object focus and classification power of R-CNN features. With the proposed contributions, this new version of Deformable ConvNets yields significant performance gains over the original model and produces leading results on the COCO benchmark for object detection and instance segmentation.
AbstractList The superior performance of Deformable Convolutional Networks arises from its ability to adapt to the geometric variations of objects. Through an examination of its adaptive behavior, we observe that while the spatial support for its neural features conforms more closely than regular ConvNets to object structure, this support may nevertheless extend well beyond the region of interest, causing features to be influenced by irrelevant image content. To address this problem, we present a reformulation of Deformable ConvNets that improves its ability to focus on pertinent image regions, through increased modeling power and stronger training. The modeling power is enhanced through a more comprehensive integration of deformable convolution within the network, and by introducing a modulation mechanism that expands the scope of deformation modeling. To effectively harness this enriched modeling capability, we guide network training via a proposed feature mimicking scheme that helps the network to learn features that reflect the object focus and classification power of R-CNN features. With the proposed contributions, this new version of Deformable ConvNets yields significant performance gains over the original model and produces leading results on the COCO benchmark for object detection and instance segmentation.
Author Zhu, Xizhou
Dai, Jifeng
Lin, Stephen
Hu, Han
Author_xml – sequence: 1
  givenname: Xizhou
  surname: Zhu
  fullname: Zhu, Xizhou
  organization: Univ. of Science and Technology of China
– sequence: 2
  givenname: Han
  surname: Hu
  fullname: Hu, Han
  organization: Microsoft Research Asia
– sequence: 3
  givenname: Stephen
  surname: Lin
  fullname: Lin, Stephen
  organization: Microsoft Research
– sequence: 4
  givenname: Jifeng
  surname: Dai
  fullname: Dai, Jifeng
  organization: Microsoft Research Asia
BookMark eNpFj8tKxEAQRVtRcByzduGmP8DErqpOutudxieMDwad7ZCJVRDJJJKOgn9vQMHVhXvgcu6h2uv6jpU6BpMBmHBWrp6XGRoImTEhpx2VBOfBoQfCQH5XzcAUlBYBwoFKYnw3xhACFMHPlLti6YdttWlZl3339chj1Cs81w_9wPofnupLHkce9JLjZzvGI7UvVRs5-cu5er25finv0sXT7X15sUgbNDSmOQDWtq7YsiAKbqyVqchFKgz4xp7y3AqIgK9dYUnEF27y82RdJYQ0Vye_uw0zrz-GZlsN32s__XTB0Q9FhUb2
ContentType Conference Proceeding
DBID 6IE
6IH
CBEJK
RIE
RIO
DOI 10.1109/CVPR.2019.00953
DatabaseName IEEE Electronic Library (IEL) Conference Proceedings
IEEE Proceedings Order Plan (POP) 1998-present by volume
IEEE Xplore All Conference Proceedings
IEEE Xplore
IEEE Proceedings Order Plans (POP) 1998-present
DatabaseTitleList
Database_xml – sequence: 1
  dbid: RIE
  name: IEEE Electronic Library (IEL)
  url: https://ieeexplore.ieee.org/
  sourceTypes: Publisher
DeliveryMethod fulltext_linktorsrc
Discipline Applied Sciences
EISBN 9781728132938
1728132932
EISSN 1063-6919
EndPage 9308
ExternalDocumentID 8953797
Genre orig-research
GroupedDBID 6IE
6IH
6IL
6IN
AAWTH
ABLEC
ADZIZ
ALMA_UNASSIGNED_HOLDINGS
BEFXN
BFFAM
BGNUA
BKEBE
BPEOZ
CBEJK
CHZPO
IEGSK
IJVOP
OCL
RIE
RIL
RIO
ID FETCH-LOGICAL-i203t-5112c4cae4ef22f2b44f2c45ffa292de83554f1ff18c7643ff8670008347af323
IEDL.DBID RIE
ISICitedReferencesCount 2132
ISICitedReferencesURI http://www.webofscience.com/api/gateway?GWVersion=2&SrcApp=Summon&SrcAuth=ProQuest&DestLinkType=CitingArticles&DestApp=WOS_CPL&KeyUT=000542649302094&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
IngestDate Wed Aug 27 07:44:55 EDT 2025
IsPeerReviewed false
IsScholarly true
Language English
LinkModel DirectLink
MergedId FETCHMERGED-LOGICAL-i203t-5112c4cae4ef22f2b44f2c45ffa292de83554f1ff18c7643ff8670008347af323
PageCount 9
ParticipantIDs ieee_primary_8953797
PublicationCentury 2000
PublicationDate 2019-June
PublicationDateYYYYMMDD 2019-06-01
PublicationDate_xml – month: 06
  year: 2019
  text: 2019-June
PublicationDecade 2010
PublicationTitle Proceedings (IEEE Computer Society Conference on Computer Vision and Pattern Recognition. Online)
PublicationTitleAbbrev CVPR
PublicationYear 2019
Publisher IEEE
Publisher_xml – name: IEEE
SSID ssj0003211698
Score 2.6645617
Snippet The superior performance of Deformable Convolutional Networks arises from its ability to adapt to the geometric variations of objects. Through an examination...
SourceID ieee
SourceType Publisher
StartPage 9300
SubjectTerms Categorization
Convolution
Convolutional neural networks
Deformable models
Deformation
Instance segmentation
Modulation
Object detection
Pattern recognition
Performance gain
Recognition: Detection
Retrieval
Training
Title Deformable ConvNets V2: More Deformable, Better Results
URI https://ieeexplore.ieee.org/document/8953797
WOSCitedRecordID wos000542649302094&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
hasFullText 1
inHoldings 1
isFullTextHit
isPrint
link http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwlV1Nb8IwDLUA7bAT22Dat3LYkY6StqTZcWxohw0htCFuqE1tCQkBoi2_f3GoYIdddoscRZEcWX52_GyARws6M6PI2jcxJcfX5PFvm5dKFZkwyozM3Et_qNEons30uAadAxcGEV3xGT7x0v3lZ2tTcqqsG-soUFrVoa6U2nO1DvmUwEYyfR1X3Xt6vu4OpuMJ125xQ0rNw49_jU9x3mPY_N-9Z9A-0vDE-OBgzqGGqwtoVrhRVFaZt0C9ooOe6RKFPbobYZGLqXwWn_YicdzsiBdH3hETzMtlkbfhe_j2NXj3qoEI3kL6QeExODKhSTBEkpJkGoZkBRFRIrXMMGbwQD2iXmyUhRpEMbNwLMoKVUKBDC6hsVqv8ApEYi056pOfyMwGhNKkNiDV3DvGHoySNLiGFuthvtn3vJhXKrj5W3wLp6zofQnVHTSKbYn3cGJ2xSLfPriH-gEfrJKG
linkProvider IEEE
linkToHtml http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwlV1LawIxEB6sLbQn22rpuzn06NbdbGI2PfYhluoiYsWb7GYzIIgWd_X3N4mL9tBLb2FCCEwY5pvJfDMAjwZ0ZkqgsW-0lBxfomd_27yUCq4YzxTN3Ev3RBxHk4kcVKC548JorV3xmX6yS_eXny3V2qbKWpHkoZDiAA45YzTYsrV2GZXQxDJtGZX9ewJftl7Hg6Gt3rItKaUdf_xrgIrzH53a_24-hcaeiEcGOxdzBhW9OIdaiRxJaZd5HcSbduAznWtijm5iXeRkTJ9J31xE9ptN8uLoO2So8_W8yBvw1XkfvXa9ciSCN6N-WHgWHimmEs00Uoo0ZQyNgCMmVNJMRxY-YIAYREoYsIEYWR6OwVlMJBjS8AKqi-VCXwJJjC3zNvoJzUxISFVqQlJpu8eYgzxJwyuoWz1Mv7ddL6alCq7_Fj_AcXfU7017H_HnDZxYpW8Lqm6hWqzW-g6O1KaY5at792g_8jKVzQ
openUrl ctx_ver=Z39.88-2004&ctx_enc=info%3Aofi%2Fenc%3AUTF-8&rfr_id=info%3Asid%2Fsummon.serialssolutions.com&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Abook&rft.genre=proceeding&rft.title=Proceedings+%28IEEE+Computer+Society+Conference+on+Computer+Vision+and+Pattern+Recognition.+Online%29&rft.atitle=Deformable+ConvNets+V2%3A+More+Deformable%2C+Better+Results&rft.au=Zhu%2C+Xizhou&rft.au=Hu%2C+Han&rft.au=Lin%2C+Stephen&rft.au=Dai%2C+Jifeng&rft.date=2019-06-01&rft.pub=IEEE&rft.eissn=1063-6919&rft.spage=9300&rft.epage=9308&rft_id=info:doi/10.1109%2FCVPR.2019.00953&rft.externalDocID=8953797