Bottleneck Transformers for Visual Recognition

We present BoTNet, a conceptually simple yet powerful backbone architecture that incorporates self-attention for multiple computer vision tasks including image classification, object detection and instance segmentation. By just replacing the spatial convolutions with global self-attention in the fin...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Proceedings (IEEE Computer Society Conference on Computer Vision and Pattern Recognition. Online) S. 16514 - 16524
Hauptverfasser: Srinivas, Aravind, Lin, Tsung-Yi, Parmar, Niki, Shlens, Jonathon, Abbeel, Pieter, Vaswani, Ashish
Format: Tagungsbericht
Sprache:Englisch
Veröffentlicht: IEEE 01.06.2021
Schlagworte:
ISSN:1063-6919
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Abstract We present BoTNet, a conceptually simple yet powerful backbone architecture that incorporates self-attention for multiple computer vision tasks including image classification, object detection and instance segmentation. By just replacing the spatial convolutions with global self-attention in the final three bottleneck blocks of a ResNet and no other changes, our approach improves upon the baselines significantly on instance segmentation and object detection while also reducing the parameters, with minimal overhead in latency. Through the design of BoTNet, we also point out how ResNet bottleneck blocks with self-attention can be viewed as Transformer blocks. Without any bells and whistles, BoTNet achieves 44.4% Mask AP and 49.7% Box AP on the COCO Instance Segmentation benchmark using the Mask R-CNN framework; surpassing the previous best published single model and single scale results of ResNeSt [67] evaluated on the COCO validation set. Finally, we present a simple adaptation of the BoTNet design for image classification, resulting in models that achieve a strong performance of 84.7% top-1 accuracy on the ImageNet benchmark while being up to 1.64x faster in "compute" 1 time than the popular EfficientNet models on TPU-v3 hardware. We hope our simple and effective approach will serve as a strong baseline for future research in self-attention models for vision. 2
AbstractList We present BoTNet, a conceptually simple yet powerful backbone architecture that incorporates self-attention for multiple computer vision tasks including image classification, object detection and instance segmentation. By just replacing the spatial convolutions with global self-attention in the final three bottleneck blocks of a ResNet and no other changes, our approach improves upon the baselines significantly on instance segmentation and object detection while also reducing the parameters, with minimal overhead in latency. Through the design of BoTNet, we also point out how ResNet bottleneck blocks with self-attention can be viewed as Transformer blocks. Without any bells and whistles, BoTNet achieves 44.4% Mask AP and 49.7% Box AP on the COCO Instance Segmentation benchmark using the Mask R-CNN framework; surpassing the previous best published single model and single scale results of ResNeSt [67] evaluated on the COCO validation set. Finally, we present a simple adaptation of the BoTNet design for image classification, resulting in models that achieve a strong performance of 84.7% top-1 accuracy on the ImageNet benchmark while being up to 1.64x faster in "compute" 1 time than the popular EfficientNet models on TPU-v3 hardware. We hope our simple and effective approach will serve as a strong baseline for future research in self-attention models for vision. 2
Author Parmar, Niki
Lin, Tsung-Yi
Vaswani, Ashish
Srinivas, Aravind
Shlens, Jonathon
Abbeel, Pieter
Author_xml – sequence: 1
  givenname: Aravind
  surname: Srinivas
  fullname: Srinivas, Aravind
  email: aravind@cs.berkeley.edu
  organization: UC Berkeley
– sequence: 2
  givenname: Tsung-Yi
  surname: Lin
  fullname: Lin, Tsung-Yi
  organization: Google Research
– sequence: 3
  givenname: Niki
  surname: Parmar
  fullname: Parmar, Niki
  organization: Google Research
– sequence: 4
  givenname: Jonathon
  surname: Shlens
  fullname: Shlens, Jonathon
  organization: Google Research
– sequence: 5
  givenname: Pieter
  surname: Abbeel
  fullname: Abbeel, Pieter
  organization: UC Berkeley
– sequence: 6
  givenname: Ashish
  surname: Vaswani
  fullname: Vaswani, Ashish
  organization: Google Research
BookMark eNotjUtOwzAUAA0Cibb0BLDIBRLe8y_2EiJ-UiVQVbqtnOQZGVIbxWHB7YkEs5ndzJKdxRSJsWuEChHsTbN_3UotRV1x4FgBaq5O2BK1VlIqsPyULRC0KLVFe8HWOX8AgOCI2poFq-7SNA0UqfssdqOL2afxSGMuZhf7kL_dUGypS-8xTCHFS3bu3ZBp_e8Ve3u43zVP5ebl8bm53ZSBg5jK1s232ljsenSIkjuUrXcSyFvTOSG9A2W89dhb4r1EZWTvDJKmVmlDYsWu_rqBiA5fYzi68edgVT2D4hepR0Xz
CODEN IEEPAD
ContentType Conference Proceeding
DBID 6IE
6IH
CBEJK
RIE
RIO
DOI 10.1109/CVPR46437.2021.01625
DatabaseName IEEE Electronic Library (IEL) Conference Proceedings
IEEE Proceedings Order Plan (POP) 1998-present by volume
IEEE Xplore All Conference Proceedings
IEEE Electronic Library (IEL)
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 1665445092
9781665445092
EISSN 1063-6919
EndPage 16524
ExternalDocumentID 9577771
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-ba6917891cd1a1142a14bfa40ef98ca34fa058f9f1d9e2d41584da81e6eb568e3
IEDL.DBID RIE
ISICitedReferencesCount 982
ISICitedReferencesURI http://www.webofscience.com/api/gateway?GWVersion=2&SrcApp=Summon&SrcAuth=ProQuest&DestLinkType=CitingArticles&DestApp=WOS_CPL&KeyUT=000742075006072&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 02:28:30 EDT 2025
IsPeerReviewed false
IsScholarly true
Language English
LinkModel DirectLink
MergedId FETCHMERGED-LOGICAL-i203t-ba6917891cd1a1142a14bfa40ef98ca34fa058f9f1d9e2d41584da81e6eb568e3
PageCount 11
ParticipantIDs ieee_primary_9577771
PublicationCentury 2000
PublicationDate 2021-June
PublicationDateYYYYMMDD 2021-06-01
PublicationDate_xml – month: 06
  year: 2021
  text: 2021-June
PublicationDecade 2020
PublicationTitle Proceedings (IEEE Computer Society Conference on Computer Vision and Pattern Recognition. Online)
PublicationTitleAbbrev CVPR
PublicationYear 2021
Publisher IEEE
Publisher_xml – name: IEEE
SSID ssj0003211698
Score 2.6880398
Snippet We present BoTNet, a conceptually simple yet powerful backbone architecture that incorporates self-attention for multiple computer vision tasks including image...
SourceID ieee
SourceType Publisher
StartPage 16514
SubjectTerms Adaptation models
Botnet
Computational modeling
Computer architecture
Computer vision
Image segmentation
Object detection
Title Bottleneck Transformers for Visual Recognition
URI https://ieeexplore.ieee.org/document/9577771
WOSCitedRecordID wos000742075006072&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/eLvHCXMwlV1NSwMxEB1q8eCpait-k4NHt9002W1ytVg8SCmlLr2VfEygKLvS3fr7Tbbb6sGLuSSEQEhCeDPJvDcADypIjKdBClELFXFqXORHphEdmYC-aEWts529jqZTsVzKWQseD1wYRKyDz7AfmvVfvi3MNjyVDWQy8sX7Oke-2nG1Du8pzHsyqRQNO47GcjDOZnMe_qW8FzikfW_bhHzYv3Ko1BAy6fxv8lPo_XDxyOyAMmfQwvwcOo3xSJqrWXah_1QEOeIczTtZ7K1Rb9sRX5NsXW7VB5nvo4WKvAdvk-fF-CVqkiFE62HMqkir1HtWQlJjqQoEWEW5dorH6KQwinGn4kQ46aiVOLQelwW3SlBMUSepQHYB7bzI8RKI8yhtGI8txoorq7U1OtHMJFYapqm7gm5Y_upzp3exalZ-_Xf3DZyE_d2FT91Cu9ps8Q6OzVe1Ljf39SF9A927k-Q
linkProvider IEEE
linkToHtml http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwlV1LTwIxEJ4QNNETKhjf9uDRhe1ud2mvEglGJIQg4Ub6mCZEsmt4-PttlwU9eLGXNk2Tpm2ab6ad7xuAB-klxlMvhai4DBjVNnAj04C2tUdfNLzQ2Z7024MBn07FsAKPey4MIhbBZ9j0zeIv3-R645_KWiJpu-J8nYOEsYhu2Vr7F5XY-TKp4CU_joai1ZkMR8z_TDk_MKJNZ934jNi_sqgUINKt_W_6E2j8sPHIcI8zp1DB7AxqpflIysu5qkPzKfeCxBnqDzLe2aPOuiOuJpP5aiMXZLSLF8qzBrx3n8edXlCmQwjmURivAyVT51txQbWh0lNgJWXKShaiFVzLmFkZJtwKS43AyDhk5sxITjFFlaQc43OoZnmGF0Csw2kds9BgKJk0ShmtEhXrxAgdK2ovoe6XP_vcKl7MypVf_d19D0e98Vt_1n8ZvF7Dsd_rbTDVDVTXyw3ewqH-Ws9Xy7viwL4Bv9qXKw
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%3Ajournal&rft.genre=proceeding&rft.title=Proceedings+%28IEEE+Computer+Society+Conference+on+Computer+Vision+and+Pattern+Recognition.+Online%29&rft.atitle=Bottleneck+Transformers+for+Visual+Recognition&rft.au=Srinivas%2C+Aravind&rft.au=Lin%2C+Tsung-Yi&rft.au=Parmar%2C+Niki&rft.au=Shlens%2C+Jonathon&rft.date=2021-06-01&rft.pub=IEEE&rft.eissn=1063-6919&rft.spage=16514&rft.epage=16524&rft_id=info:doi/10.1109%2FCVPR46437.2021.01625&rft.externalDocID=9577771