Distribution Shift Inversion for Out-of-Distribution Prediction

Machine learning society has witnessed the emergence of a myriad of Out-of-Distribution (OoD) algorithms, which address the distribution shift between the training and the testing distribution by searching for a unified predictor or invariant feature representation. However, the task of directly mit...

Full description

Saved in:
Bibliographic Details
Published in:Proceedings (IEEE Computer Society Conference on Computer Vision and Pattern Recognition. Online) pp. 3592 - 3602
Main Authors: Yu, Runpeng, Liu, Songhua, Yang, Xingyi, Wang, Xinchao
Format: Conference Proceeding
Language:English
Published: IEEE 01.06.2023
Subjects:
ISSN:1063-6919
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Abstract Machine learning society has witnessed the emergence of a myriad of Out-of-Distribution (OoD) algorithms, which address the distribution shift between the training and the testing distribution by searching for a unified predictor or invariant feature representation. However, the task of directly mitigating the distribution shift in the unseen testing set is rarely investigated, due to the unavailability of the testing distribution during the training phase and thus the impossibility of training a distribution translator mapping between the training and testing distribution. In this paper, we explore how to bypass the requirement of testing distribution for distribution translator training and make the distribution translation useful for OoD prediction. We propose a portable Distribution Shift Inversion (DSI) algorithm, in which, before being fed into the prediction model, the OoD testing samples are first linearly combined with additional Gaussian noise and then transferred back towards the training distribution using a diffusion model trained only on the source distribution. Theoretical analysis reveals the feasibility of our method. Experimental results, on both multiple-domain generalization datasets and single-domain generalization datasets, show that our method provides a general performance gain when plugged into a wide range of commonly used OoD algorithms. Our code is available at https://github.com/yu-rp/Distribution-Shift-Iverson.
AbstractList Machine learning society has witnessed the emergence of a myriad of Out-of-Distribution (OoD) algorithms, which address the distribution shift between the training and the testing distribution by searching for a unified predictor or invariant feature representation. However, the task of directly mitigating the distribution shift in the unseen testing set is rarely investigated, due to the unavailability of the testing distribution during the training phase and thus the impossibility of training a distribution translator mapping between the training and testing distribution. In this paper, we explore how to bypass the requirement of testing distribution for distribution translator training and make the distribution translation useful for OoD prediction. We propose a portable Distribution Shift Inversion (DSI) algorithm, in which, before being fed into the prediction model, the OoD testing samples are first linearly combined with additional Gaussian noise and then transferred back towards the training distribution using a diffusion model trained only on the source distribution. Theoretical analysis reveals the feasibility of our method. Experimental results, on both multiple-domain generalization datasets and single-domain generalization datasets, show that our method provides a general performance gain when plugged into a wide range of commonly used OoD algorithms. Our code is available at https://github.com/yu-rp/Distribution-Shift-Iverson.
Author Liu, Songhua
Yu, Runpeng
Yang, Xingyi
Wang, Xinchao
Author_xml – sequence: 1
  givenname: Runpeng
  surname: Yu
  fullname: Yu, Runpeng
  email: r.yu@u.nus.edu
  organization: National University of Singapore
– sequence: 2
  givenname: Songhua
  surname: Liu
  fullname: Liu, Songhua
  email: songhua.liu@u.nus.edu
  organization: National University of Singapore
– sequence: 3
  givenname: Xingyi
  surname: Yang
  fullname: Yang, Xingyi
  email: xyang@u.nus.edu
  organization: National University of Singapore
– sequence: 4
  givenname: Xinchao
  surname: Wang
  fullname: Wang, Xinchao
  email: xinchao@nus.edu.sg
  organization: National University of Singapore
BookMark eNpVj91Kw0AUhFdRsNa8QS_yAhvP2ZNsslci0Wqh0OLfbcluzuKKJrJJBd_eFL3xambgY5g5Fydd37EQC4QMEcxl_bJ9KFSpTKZAUQZABRyJxJSmmhwBKlMdixmCJqkNmjORDMMbTJxC1KaaiaubMIwx2P0Y-i59fA1-TFfdF8fhkH0f081-lL2X_7ht5Da4g70Qp755Hzj507l4Xt4-1fdyvblb1ddrGRTko3TOAkKbtxpNA2xL64rG2QqQi0p7dJoKbpWzuSNFmoCniUpPr7xm2xqai8Vvb2Dm3WcMH0383iEoIINEP3HiTOU
CODEN IEEPAD
ContentType Conference Proceeding
DBID 6IE
6IH
CBEJK
RIE
RIO
DOI 10.1109/CVPR52729.2023.00350
DatabaseName IEEE Electronic Library (IEL) Conference Proceedings
IEEE Proceedings Order Plan (POP) 1998-present by volume
IEEE Xplore All Conference Proceedings
IEEE/IET Electronic Library
IEEE Proceedings Order Plans (POP) 1998-present
DatabaseTitleList
Database_xml – sequence: 1
  dbid: RIE
  name: IEEE/IET Electronic Library
  url: https://ieeexplore.ieee.org/
  sourceTypes: Publisher
DeliveryMethod fulltext_linktorsrc
Discipline Applied Sciences
EISBN 9798350301298
EISSN 1063-6919
EndPage 3602
ExternalDocumentID 10203913
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-i204t-ccb010d4d619a0eb7bc5acb801e586f1c635ed2cb4c323630e03226202f6ebd93
IEDL.DBID RIE
ISICitedReferencesCount 6
ISICitedReferencesURI http://www.webofscience.com/api/gateway?GWVersion=2&SrcApp=Summon&SrcAuth=ProQuest&DestLinkType=CitingArticles&DestApp=WOS_CPL&KeyUT=001058542603085&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 02:56:29 EDT 2025
IsPeerReviewed false
IsScholarly true
Language English
LinkModel DirectLink
MergedId FETCHMERGED-LOGICAL-i204t-ccb010d4d619a0eb7bc5acb801e586f1c635ed2cb4c323630e03226202f6ebd93
PageCount 11
ParticipantIDs ieee_primary_10203913
PublicationCentury 2000
PublicationDate 2023-June
PublicationDateYYYYMMDD 2023-06-01
PublicationDate_xml – month: 06
  year: 2023
  text: 2023-June
PublicationDecade 2020
PublicationTitle Proceedings (IEEE Computer Society Conference on Computer Vision and Pattern Recognition. Online)
PublicationTitleAbbrev CVPR
PublicationYear 2023
Publisher IEEE
Publisher_xml – name: IEEE
SSID ssj0003211698
Score 2.3811908
Snippet Machine learning society has witnessed the emergence of a myriad of Out-of-Distribution (OoD) algorithms, which address the distribution shift between the...
SourceID ieee
SourceType Publisher
StartPage 3592
SubjectTerms continual
Gaussian noise
low-shot
Machine learning
Machine learning algorithms
meta
or long-tail learning
Performance gain
Prediction algorithms
Predictive models
Training
Transfer
Title Distribution Shift Inversion for Out-of-Distribution Prediction
URI https://ieeexplore.ieee.org/document/10203913
WOSCitedRecordID wos001058542603085&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/eLvHCXMwlV25TgMxEB1BREEVjiBuuaB18Np7eCuKQEQVVlxKF63tsUiToGTD9zN2lqAUFGxlrSxZGsuaN8d7A3AjbJo6JxUnrICcPL7kdeh1TbQXiS6lV9rHYRPFaKTH47JqyeqRC4OIsfkM-2EZa_lublchVUYvXAZBc7ULu0WRr8lam4SKolAmL3VLj0tEeTt4r54zSeixH2aE92MRbWuISvQhw-4_Tz-A3i8bj1UbP3MIOzg7gm4LH1n7OJfHcHcfRHDb-VXs5WPqGxZkNGJCjBE4ZU-rhs8939pXLUKpJix78DZ8eB088nY-Ap9KkTbcWkPRlEsdBUG1QFMYm9XWkM_BTOc-sQQm0ElrUqukypVAMlIQoJc-R-NKdQKd2XyGp8DQZfTVxtcEL6QWxgZglRYSramlUmfQCwaZfK4lMCY_tjj_4_8F7Aebr3uqLqHTLFZ4BXv2q5kuF9fx4r4BnC6Y3g
linkProvider IEEE
linkToHtml http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwlV1LTwIxEJ4omugJHxjf7sFrsdt2l-7JA0owIm4UDTeybaeRCxhY_P22ZcVw8OCemk2TJtM0883j-wbgmmohjGGcOKyAxHl8Rgrf6xpLS2OZMculDcMmWv2-HA6zvCKrBy4MIobmM2z6Zajlm6le-FSZe-HMC5rzTdhKhGB0SddapVS4C2bSTFYEuZhmN-33_CVhDj82_ZTwZiijrY1RCV6kU__n-XvQ-OXjRfnK0-zDBk4OoF4ByKh6nvNDuL3zMrjVBKvo9WNsy8gLaYSUWOTgafS8KMnUkrV9-cwXa_yyAW-d-0G7S6oJCWTMqCiJ1srFU0YYFwYVFFVL6aTQynkdTGRqY-3gBBqmldCc8ZRTdEbyEvTMpqhMxo-gNplO8BgiNIn7CmULBzCYpEp7aCVaDLUqGOcn0PAGGX0uRTBGP7Y4_eP_Fex0B0-9Ue-h_3gGu97-yw6rc6iVswVewLb-Ksfz2WW4xG_AWZwl
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=Distribution+Shift+Inversion+for+Out-of-Distribution+Prediction&rft.au=Yu%2C+Runpeng&rft.au=Liu%2C+Songhua&rft.au=Yang%2C+Xingyi&rft.au=Wang%2C+Xinchao&rft.date=2023-06-01&rft.pub=IEEE&rft.eissn=1063-6919&rft.spage=3592&rft.epage=3602&rft_id=info:doi/10.1109%2FCVPR52729.2023.00350&rft.externalDocID=10203913