DS-UI: Dual-Supervised Mixture of Gaussian Mixture Models for Uncertainty Inference in Image Recognition

This paper proposes a dual-supervised uncertainty inference (DS-UI) framework for improving Bayesian estimation-based UI in DNN-based image recognition. In the DS-UI, we combine the classifier of a DNN, i.e. , the last fully-connected (FC) layer, with a mixture of Gaussian mixture models (MoGMM) to...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:IEEE transactions on image processing Jg. 30; S. 9208 - 9219
Hauptverfasser: Xie, Jiyang, Ma, Zhanyu, Xue, Jing-Hao, Zhang, Guoqiang, Sun, Jian, Zheng, Yinhe, Guo, Jun
Format: Journal Article
Sprache:Englisch
Veröffentlicht: New York IEEE 2021
The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
Schlagworte:
ISSN:1057-7149, 1941-0042, 1941-0042
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Abstract This paper proposes a dual-supervised uncertainty inference (DS-UI) framework for improving Bayesian estimation-based UI in DNN-based image recognition. In the DS-UI, we combine the classifier of a DNN, i.e. , the last fully-connected (FC) layer, with a mixture of Gaussian mixture models (MoGMM) to obtain an MoGMM-FC layer. Unlike existing UI methods for DNNs, which only calculate the means or modes of the DNN outputs' distributions, the proposed MoGMM-FC layer acts as a probabilistic interpreter for the features that are inputs of the classifier to directly calculate the probabilities of them for the DS-UI. In addition, we propose a dual-supervised stochastic gradient-based variational Bayes (DS-SGVB) algorithm for the MoGMM-FC layer optimization. Unlike conventional SGVB and optimization algorithms in other UI methods, the DS-SGVB not only models the samples in the specific class for each Gaussian mixture model (GMM) in the MoGMM, but also considers the negative samples from other classes for the GMM to reduce the intra-class distances and enlarge the inter-class margins simultaneously for enhancing the learning ability of the MoGMM-FC layer in the DS-UI. Experimental results show the DS-UI outperforms the state-of-the-art UI methods in misclassification detection. We further evaluate the DS-UI in open-set out-of-domain/-distribution detection and find statistically significant improvements. Visualizations of the feature spaces demonstrate the superiority of the DS-UI. Codes are available at https://github.com/PRIS-CV/DS-UI .
AbstractList This paper proposes a dual-supervised uncertainty inference (DS-UI) framework for improving Bayesian estimation-based UI in DNN-based image recognition. In the DS-UI, we combine the classifier of a DNN, i.e. , the last fully-connected (FC) layer, with a mixture of Gaussian mixture models (MoGMM) to obtain an MoGMM-FC layer. Unlike existing UI methods for DNNs, which only calculate the means or modes of the DNN outputs' distributions, the proposed MoGMM-FC layer acts as a probabilistic interpreter for the features that are inputs of the classifier to directly calculate the probabilities of them for the DS-UI. In addition, we propose a dual-supervised stochastic gradient-based variational Bayes (DS-SGVB) algorithm for the MoGMM-FC layer optimization. Unlike conventional SGVB and optimization algorithms in other UI methods, the DS-SGVB not only models the samples in the specific class for each Gaussian mixture model (GMM) in the MoGMM, but also considers the negative samples from other classes for the GMM to reduce the intra-class distances and enlarge the inter-class margins simultaneously for enhancing the learning ability of the MoGMM-FC layer in the DS-UI. Experimental results show the DS-UI outperforms the state-of-the-art UI methods in misclassification detection. We further evaluate the DS-UI in open-set out-of-domain/-distribution detection and find statistically significant improvements. Visualizations of the feature spaces demonstrate the superiority of the DS-UI. Codes are available at https://github.com/PRIS-CV/DS-UI .
This paper proposes a dual-supervised uncertainty inference (DS-UI) framework for improving Bayesian estimation-based UI in DNN-based image recognition. In the DS-UI, we combine the classifier of a DNN, i.e., the last fully-connected (FC) layer, with a mixture of Gaussian mixture models (MoGMM) to obtain an MoGMM-FC layer. Unlike existing UI methods for DNNs, which only calculate the means or modes of the DNN outputs' distributions, the proposed MoGMM-FC layer acts as a probabilistic interpreter for the features that are inputs of the classifier to directly calculate the probabilities of them for the DS-UI. In addition, we propose a dual-supervised stochastic gradient-based variational Bayes (DS-SGVB) algorithm for the MoGMM-FC layer optimization. Unlike conventional SGVB and optimization algorithms in other UI methods, the DS-SGVB not only models the samples in the specific class for each Gaussian mixture model (GMM) in the MoGMM, but also considers the negative samples from other classes for the GMM to reduce the intra-class distances and enlarge the inter-class margins simultaneously for enhancing the learning ability of the MoGMM-FC layer in the DS-UI. Experimental results show the DS-UI outperforms the state-of-the-art UI methods in misclassification detection. We further evaluate the DS-UI in open-set out-of-domain/-distribution detection and find statistically significant improvements. Visualizations of the feature spaces demonstrate the superiority of the DS-UI. Codes are available at https://github.com/PRIS-CV/DS-UI.This paper proposes a dual-supervised uncertainty inference (DS-UI) framework for improving Bayesian estimation-based UI in DNN-based image recognition. In the DS-UI, we combine the classifier of a DNN, i.e., the last fully-connected (FC) layer, with a mixture of Gaussian mixture models (MoGMM) to obtain an MoGMM-FC layer. Unlike existing UI methods for DNNs, which only calculate the means or modes of the DNN outputs' distributions, the proposed MoGMM-FC layer acts as a probabilistic interpreter for the features that are inputs of the classifier to directly calculate the probabilities of them for the DS-UI. In addition, we propose a dual-supervised stochastic gradient-based variational Bayes (DS-SGVB) algorithm for the MoGMM-FC layer optimization. Unlike conventional SGVB and optimization algorithms in other UI methods, the DS-SGVB not only models the samples in the specific class for each Gaussian mixture model (GMM) in the MoGMM, but also considers the negative samples from other classes for the GMM to reduce the intra-class distances and enlarge the inter-class margins simultaneously for enhancing the learning ability of the MoGMM-FC layer in the DS-UI. Experimental results show the DS-UI outperforms the state-of-the-art UI methods in misclassification detection. We further evaluate the DS-UI in open-set out-of-domain/-distribution detection and find statistically significant improvements. Visualizations of the feature spaces demonstrate the superiority of the DS-UI. Codes are available at https://github.com/PRIS-CV/DS-UI.
Author Xue, Jing-Hao
Ma, Zhanyu
Xie, Jiyang
Zhang, Guoqiang
Zheng, Yinhe
Guo, Jun
Sun, Jian
Author_xml – sequence: 1
  givenname: Jiyang
  orcidid: 0000-0003-3659-9476
  surname: Xie
  fullname: Xie, Jiyang
  email: xiejiyang2013@bupt.edu.cn
  organization: Pattern Recognition and Intelligent Systems Laboratory, Beijing University of Posts and Telecommunications, Beijing, China
– sequence: 2
  givenname: Zhanyu
  orcidid: 0000-0003-2950-2488
  surname: Ma
  fullname: Ma, Zhanyu
  email: mazhanyu@bupt.edu.cn
  organization: Pattern Recognition and Intelligent Systems Laboratory, Beijing University of Posts and Telecommunications, Beijing, China
– sequence: 3
  givenname: Jing-Hao
  orcidid: 0000-0003-1174-610X
  surname: Xue
  fullname: Xue, Jing-Hao
  email: jinghao.xue@ucl.ac.uk
  organization: Department of Statistical Science, University College London, London, U.K
– sequence: 4
  givenname: Guoqiang
  orcidid: 0000-0003-4521-542X
  surname: Zhang
  fullname: Zhang, Guoqiang
  email: guoqiang.zhang@uts.edu.au
  organization: School of Electrical and Data Engineering, University of Technology Sydney, Sydney, NSW, Australia
– sequence: 5
  givenname: Jian
  surname: Sun
  fullname: Sun, Jian
  email: jiansun_china@hotmail.com
  organization: Pattern Recognition and Intelligent Systems Laboratory, Beijing University of Posts and Telecommunications, Beijing, China
– sequence: 6
  givenname: Yinhe
  orcidid: 0000-0002-7029-5671
  surname: Zheng
  fullname: Zheng, Yinhe
  email: zhengyinhe1@163.com
  organization: Department of Computer Science and Technology, Tsinghua University, Beijing, China
– sequence: 7
  givenname: Jun
  orcidid: 0000-0001-9045-1339
  surname: Guo
  fullname: Guo, Jun
  email: guojun@bupt.edu.cn
  organization: Pattern Recognition and Intelligent Systems Laboratory, Beijing University of Posts and Telecommunications, Beijing, China
BookMark eNp9kc1P3DAQxS1EVb56R-rFEpdesrXHdhJzQ9DSSKBWwJ4jrz2mRll7sZOq_PfNahEHDj3N6On3Zkbzjsh-TBEJOeVswTnTXx-6XwtgwBeCg1BK7ZFDriWvGJOwP_dMNVXDpT4gR6U8Mcal4vVHciBkI7Ro6kPy--q-Wnbn9GoyQ3U_bTD_CQUdvQ1_xykjTZ5em6mUYOKbdpscDoX6lOkyWsyjCXF8oV30mHEWaIi0W5tHpHdo02MMY0jxhHzwZij46bUek-X3bw-XP6qbn9fd5cVNZQXIsRIro1ZeOMHtynrUljm3Eq0FNNJ756T0LXjrGg2AjQFntQPJhRGSKdtqcUy-7OZucnqesIz9OhSLw2Aipqn0oLQEzWtoZ_TsHfqUphzn67aUangrGjZTbEfZnErJ6PtNDmuTX3rO-m0K_ZxCv02hf01httTvLDaMZvuFMZsw_M_4eWcMiPi2R9dMAYD4B9dilQU
CODEN IIPRE4
CitedBy_id crossref_primary_10_1016_j_eswa_2023_122295
crossref_primary_10_1016_j_eswa_2024_124932
crossref_primary_10_1109_TIP_2023_3263099
crossref_primary_10_1109_TIM_2025_3529072
Cites_doi 10.1109/CVPR.2019.00241
10.1109/CVPR.2016.90
10.1080/10618600.2016.1200472
10.1109/TIP.2020.2970248
10.1109/TPAMI.2011.63
10.1109/ICASSP.2015.7178776
10.1109/CVPR.2018.00169
10.1007/s11263-015-0816-y
10.1109/CVPR42600.2020.01183
10.1109/CVPR42600.2020.00977
10.1109/CVPR42600.2020.01106
10.5244/C.31.42
10.1109/TNNLS.2020.2980004
10.1109/TIP.2020.2973812
10.1016/j.csda.2006.01.001
10.1109/CVPR42600.2020.00575
10.1109/CVPR46437.2021.01131
10.1109/TPAMI.2019.2956930
10.1109/TIP.2021.3055617
ContentType Journal Article
Copyright Copyright The Institute of Electrical and Electronics Engineers, Inc. (IEEE) 2021
Copyright_xml – notice: Copyright The Institute of Electrical and Electronics Engineers, Inc. (IEEE) 2021
DBID 97E
RIA
RIE
AAYXX
CITATION
7SC
7SP
8FD
JQ2
L7M
L~C
L~D
7X8
DOI 10.1109/TIP.2021.3123555
DatabaseName IEEE All-Society Periodicals Package (ASPP) 2005–Present
IEEE All-Society Periodicals Package (ASPP) 1998–Present
IEEE Electronic Library (IEL)
CrossRef
Computer and Information Systems Abstracts
Electronics & Communications Abstracts
Technology Research Database
ProQuest Computer Science Collection
Advanced Technologies Database with Aerospace
Computer and Information Systems Abstracts – Academic
Computer and Information Systems Abstracts Professional
MEDLINE - Academic
DatabaseTitle CrossRef
Technology Research Database
Computer and Information Systems Abstracts – Academic
Electronics & Communications Abstracts
ProQuest Computer Science Collection
Computer and Information Systems Abstracts
Advanced Technologies Database with Aerospace
Computer and Information Systems Abstracts Professional
MEDLINE - Academic
DatabaseTitleList
MEDLINE - Academic
Technology Research Database
Database_xml – sequence: 1
  dbid: RIE
  name: IEEE Xplore
  url: https://ieeexplore.ieee.org/
  sourceTypes: Publisher
– sequence: 2
  dbid: 7X8
  name: MEDLINE - Academic
  url: https://search.proquest.com/medline
  sourceTypes: Aggregation Database
DeliveryMethod fulltext_linktorsrc
Discipline Applied Sciences
Engineering
EISSN 1941-0042
EndPage 9219
ExternalDocumentID 10_1109_TIP_2021_3123555
9605222
Genre orig-research
GrantInformation_xml – fundername: National Key Research and Development Program of China
  grantid: 2019YFF0303300
  funderid: 10.13039/501100012166
– fundername: Subject II
  grantid: 2019YFF0303302
– fundername: National Natural Science Foundation of China (NSFC)
  grantid: 61922015; 61773071; U19B2036
  funderid: 10.13039/501100001809
– fundername: Beijing Natural Science Foundation Project
  grantid: Z200002
  funderid: 10.13039/501100004826
GroupedDBID ---
-~X
.DC
0R~
29I
4.4
53G
5GY
5VS
6IK
97E
AAJGR
AARMG
AASAJ
AAWTH
ABAZT
ABFSI
ABQJQ
ABVLG
ACGFO
ACGFS
ACIWK
AENEX
AETIX
AGQYO
AGSQL
AHBIQ
AI.
AIBXA
AKJIK
AKQYR
ALLEH
ALMA_UNASSIGNED_HOLDINGS
ASUFR
ATWAV
BEFXN
BFFAM
BGNUA
BKEBE
BPEOZ
CS3
DU5
E.L
EBS
EJD
F5P
HZ~
H~9
ICLAB
IFIPE
IFJZH
IPLJI
JAVBF
LAI
M43
MS~
O9-
OCL
P2P
RIA
RIE
RNS
TAE
TN5
VH1
AAYXX
CITATION
7SC
7SP
8FD
JQ2
L7M
L~C
L~D
7X8
ID FETCH-LOGICAL-c324t-3ba5bf3d31cbcfe9c0ddb38c2ea4ffdd44f82fcd7922e7a2dc9d2413a3405c893
IEDL.DBID RIE
ISICitedReferencesCount 9
ISICitedReferencesURI http://www.webofscience.com/api/gateway?GWVersion=2&SrcApp=Summon&SrcAuth=ProQuest&DestLinkType=CitingArticles&DestApp=WOS_CPL&KeyUT=000716696700007&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
ISSN 1057-7149
1941-0042
IngestDate Sat Sep 27 21:06:24 EDT 2025
Mon Jun 30 10:18:19 EDT 2025
Sat Nov 29 03:21:15 EST 2025
Tue Nov 18 22:33:19 EST 2025
Wed Aug 27 02:26:43 EDT 2025
IsPeerReviewed true
IsScholarly true
Language English
License https://ieeexplore.ieee.org/Xplorehelp/downloads/license-information/IEEE.html
https://doi.org/10.15223/policy-029
https://doi.org/10.15223/policy-037
LinkModel DirectLink
MergedId FETCHMERGED-LOGICAL-c324t-3ba5bf3d31cbcfe9c0ddb38c2ea4ffdd44f82fcd7922e7a2dc9d2413a3405c893
Notes ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
content type line 23
ORCID 0000-0003-3659-9476
0000-0002-7029-5671
0000-0003-4521-542X
0000-0003-2950-2488
0000-0003-1174-610X
0000-0001-9045-1339
PMID 34739376
PQID 2595718370
PQPubID 85429
PageCount 12
ParticipantIDs crossref_primary_10_1109_TIP_2021_3123555
crossref_citationtrail_10_1109_TIP_2021_3123555
proquest_miscellaneous_2594291628
proquest_journals_2595718370
ieee_primary_9605222
PublicationCentury 2000
PublicationDate 20210000
2021-00-00
20210101
PublicationDateYYYYMMDD 2021-01-01
PublicationDate_xml – year: 2021
  text: 20210000
PublicationDecade 2020
PublicationPlace New York
PublicationPlace_xml – name: New York
PublicationTitle IEEE transactions on image processing
PublicationTitleAbbrev TIP
PublicationYear 2021
Publisher IEEE
The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
Publisher_xml – name: IEEE
– name: The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
References van amersfoort (ref29) 2020
ref14
krizhevsky (ref43) 2009
simonyan (ref41) 2014
hendrycks (ref11) 2017
ref17
ref16
ref18
goodfellow (ref50) 2015
ref46
malinin (ref9) 2018
ref45
ref47
ref42
hoffman (ref36) 2013; 14
fan (ref35) 2015
kingma (ref48) 2015
malinin (ref13) 2019
ref7
ref4
ref3
ref6
altosaar (ref34) 2018
ref5
mo?ejko (ref28) 2018
piergiovanni (ref31) 2019
ref37
li (ref23) 2016
wang (ref33) 2018
ref32
netzer (ref44) 2011
ref2
teye (ref40) 2018
kong (ref15) 2020
ref1
hinton (ref22) 2012
maddox (ref12) 2019
maas (ref49) 2013; 30
bishop (ref39) 2006
lee (ref24) 2018
blundell (ref21) 2015
gal (ref10) 2016
lakshminarayanan (ref8) 2017
ranganath (ref38) 2014
ref25
ma (ref19) 2011; 33
havasi (ref30) 2021
ma (ref20) 2020; 31
antorán (ref26) 2021
guo (ref27) 2017
van d maaten (ref51) 2008; 9
References_xml – start-page: 1387
  year: 2015
  ident: ref35
  article-title: Fast second-order stochastic backpropagation for variational inference
  publication-title: Proc Adv Neural Inf Process Syst
– start-page: 1
  year: 2014
  ident: ref41
  article-title: Very deep convolutional networks for large-scale image recognition
  publication-title: Proc IEEE/CVF Conf Comput Vis Pattern Recognit (CVPR)
– ident: ref47
  doi: 10.1109/CVPR.2019.00241
– start-page: 7047
  year: 2018
  ident: ref9
  article-title: Predictive uncertainty estimation via prior networks
  publication-title: Proc Adv Neural Inf Process Syst
– start-page: 6402
  year: 2017
  ident: ref8
  article-title: Simple and scalable predictive uncertainty estimation using deep ensembles
  publication-title: Proc Adv Neural Inf Process Syst (NIPS)
– volume: 30
  start-page: 3
  year: 2013
  ident: ref49
  article-title: Rectifier nonlinearities improve neural network acoustic models
  publication-title: Proc Int Conf Mach Learn (ICML)
– start-page: 1
  year: 2021
  ident: ref26
  article-title: Getting a CLUE: A method for explaining uncertainty estimates
  publication-title: Proc Int Conf Learn Represent
– ident: ref42
  doi: 10.1109/CVPR.2016.90
– volume: 14
  start-page: 1303
  year: 2013
  ident: ref36
  article-title: Stochastic variational inference
  publication-title: J Mach Learn Res
– start-page: 4907
  year: 2018
  ident: ref40
  article-title: Bayesian uncertainty estimation for batch normalized deep networks
  publication-title: Proc Int Conf Mach Learn (ICML)
– ident: ref17
  doi: 10.1080/10618600.2016.1200472
– ident: ref6
  doi: 10.1109/TIP.2020.2970248
– volume: 33
  start-page: 2160
  year: 2011
  ident: ref19
  article-title: Bayesian estimation of beta mixture models with variational inference
  publication-title: IEEE Trans Pattern Anal Mach Intell
  doi: 10.1109/TPAMI.2011.63
– year: 2009
  ident: ref43
  article-title: Learning multiple layers of features from tiny images
– ident: ref32
  doi: 10.1109/ICASSP.2015.7178776
– ident: ref37
  doi: 10.1109/CVPR.2018.00169
– start-page: 1613
  year: 2015
  ident: ref21
  article-title: Weight uncertainty in neural networks
  publication-title: Proc Int Conf Mach Learn (ICML)
– year: 2018
  ident: ref28
  article-title: Inhibited softmax for uncertainty estimation in neural networks
  publication-title: arXiv 1810 01861
– ident: ref45
  doi: 10.1007/s11263-015-0816-y
– ident: ref16
  doi: 10.1109/CVPR42600.2020.01183
– ident: ref2
  doi: 10.1109/CVPR42600.2020.00977
– start-page: 1
  year: 2015
  ident: ref50
  article-title: Explaining and harnessing adversarial examples
  publication-title: Proc Int Conf Learn Represent
– start-page: 1050
  year: 2016
  ident: ref10
  article-title: Dropout as a Bayesian approximation: Representing model uncertainty in deep learning
  publication-title: Proc Int Conf Mach Learn (ICML)
– ident: ref1
  doi: 10.1109/CVPR42600.2020.01106
– start-page: 1961
  year: 2018
  ident: ref34
  article-title: Proximity variational inference
  publication-title: Proc Int Conf Artif Intell Statist
– start-page: 814
  year: 2014
  ident: ref38
  article-title: Black box variational inference
  publication-title: Proc Int Conf Artif Intell Statist
– ident: ref46
  doi: 10.5244/C.31.42
– start-page: 1
  year: 2015
  ident: ref48
  article-title: Adam: A method for stochastic optimization
  publication-title: Proc Int Conf Learn Represent (ICLR)
– start-page: 5152
  year: 2019
  ident: ref31
  article-title: Temporal Gaussian mixture layer for videos
  publication-title: Proc Int Conf Mach Learn (ICML)
– year: 2012
  ident: ref22
  article-title: Improving neural networks by preventing co-adaptation of feature detectors
  publication-title: arXiv 1207 0580
– ident: ref25
  doi: 10.1109/TNNLS.2020.2980004
– ident: ref3
  doi: 10.1109/TIP.2020.2973812
– start-page: 1
  year: 2018
  ident: ref24
  article-title: A simple unified framework for detecting out-of-distribution samples and adversarial attacks
  publication-title: Proc Adv Neural Inf Process Syst
– start-page: 1
  year: 2011
  ident: ref44
  article-title: Reading digits in natural images with unsupervised feature learning
  publication-title: Proc NIPS Workshop Deep Learn Unsupervised Feature Learn
– start-page: 1
  year: 2020
  ident: ref15
  article-title: SDE-Net: Equipping deep neural networks with uncertainty estimates
  publication-title: Proc Int Conf Mach Learn (ICML)
– ident: ref18
  doi: 10.1016/j.csda.2006.01.001
– ident: ref14
  doi: 10.1109/CVPR42600.2020.00575
– start-page: 1788
  year: 2016
  ident: ref23
  article-title: Preconditioned stochastic gradient Langevin dynamics for deep neural networks
  publication-title: Proc AAAI Conf Artif Intell
– start-page: 1
  year: 2021
  ident: ref30
  article-title: Training independent subnetworks for robust prediction
  publication-title: Proc Int Conf Learn Represent
– ident: ref5
  doi: 10.1109/CVPR46437.2021.01131
– ident: ref7
  doi: 10.1109/TPAMI.2019.2956930
– volume: 9
  start-page: 2579
  year: 2008
  ident: ref51
  article-title: Visualizing data using t-SNE
  publication-title: J Mach Learn Res
– volume: 31
  start-page: 2240
  year: 2020
  ident: ref20
  article-title: Insights into multiple/single lower bound approximation for extended variational inference in non-Gaussian structured data modeling
  publication-title: IEEE Trans Neural Netw Learn Syst
– start-page: 14520
  year: 2019
  ident: ref13
  article-title: Reverse KL-divergence training of prior networks: Improved uncertainty and adversarial robustness
  publication-title: Proc Adv Neural Inf Process Syst
– start-page: 13153
  year: 2019
  ident: ref12
  article-title: A simple baseline for Bayesian uncertainty in deep learning
  publication-title: Proc Adv Neural Inf Process Syst
– year: 2006
  ident: ref39
  publication-title: Pattern Recognition and Machine Learning
– start-page: 1
  year: 2017
  ident: ref11
  article-title: A baseline for detecting misclassified and out-of-distribution examples in neural networks
  publication-title: Proc Int Conf Learn Represent (ICLR)
– ident: ref4
  doi: 10.1109/TIP.2021.3055617
– start-page: 1321
  year: 2017
  ident: ref27
  article-title: On calibration of modern neural networks
  publication-title: Proc Int Conf Mach Learn (ICML)
– start-page: 249
  year: 2018
  ident: ref33
  article-title: An unsupervised deep learning framework via integrated optimization of representation learning and GMM-based modeling
  publication-title: Proc Asian Conf Comput Vis
– start-page: 9690
  year: 2020
  ident: ref29
  article-title: Uncertainty estimation using a single deep deterministic neural network
  publication-title: Proc Int Conf Mach Learn
SSID ssj0014516
Score 2.427453
Snippet This paper proposes a dual-supervised uncertainty inference (DS-UI) framework for improving Bayesian estimation-based UI in DNN-based image recognition. In the...
SourceID proquest
crossref
ieee
SourceType Aggregation Database
Enrichment Source
Index Database
Publisher
StartPage 9208
SubjectTerms Algorithms
Bayes methods
Classifiers
Deep learning
dual supervised framework
Feature extraction
Image recognition
Inference
mixture of gaussian mixture models
Object recognition
Optimization
Probabilistic logic
Probabilistic models
Statistical methods
Stochastic processes
Uncertainty
uncertainty inference
Title DS-UI: Dual-Supervised Mixture of Gaussian Mixture Models for Uncertainty Inference in Image Recognition
URI https://ieeexplore.ieee.org/document/9605222
https://www.proquest.com/docview/2595718370
https://www.proquest.com/docview/2594291628
Volume 30
WOSCitedRecordID wos000716696700007&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
journalDatabaseRights – providerCode: PRVIEE
  databaseName: IEEE Xplore
  customDbUrl:
  eissn: 1941-0042
  dateEnd: 99991231
  omitProxy: false
  ssIdentifier: ssj0014516
  issn: 1057-7149
  databaseCode: RIE
  dateStart: 19920101
  isFulltext: true
  titleUrlDefault: https://ieeexplore.ieee.org/
  providerName: IEEE
link http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwlV1La9wwEB7S0EN6aNo8yKZpUKGXQpS1JduyeitN0-6hISRZ2JuR9SALW3uJ1yH59x1pbZPSUOjN2OMHntc3Gs0MwEenhU0UWj-nUk4TyQQtrcioForzEhXKaBeGTYiLi3w2k5cbcDLUwlhrw-Yze-oPQy7f1Lr1S2VjRNsIF9DgvhAiW9dqDRkDP3A2ZDZTQQXC_j4lGcnxzeQSA0EWY3zK0L36YTU8CZ3gsj-8URiv8pdNDo7mfPv_PvENvO4AJfmyloC3sGGrHdjuwCXpVLfZgVdPOg_uwu3ZNZ1OPpOzVi3odbv0FqNB-p_zB59SILUj31Xb-ArL4Zwfm7ZoCKJcMsVnhq0Eq0cy6UsGybwik19on8hVvyuprvZgev7t5usP2g1doBqx1Yoig9LSccNjXWpnpY6MKXmumVWJc8YkicuZ00ZIxqxQzGhpfG5OcYR-GtHPPmxWdWUPgEjjjMg5ty7mKA-pVNaUMkKMFmd5JNkIxv3PL3TXkdwPxlgUITKJZIGcKzznio5zI_g03LFcd-P4B-2uZ89A13FmBEc9f4tOXZsCY8AUnTQX0Qg-DJdR0Xz2RFW2bgMN-u44Y_nh809-B1v-_ev1mSPYXN219j281PereXN3jDI7y4-DzP4G4l_nuw
linkProvider IEEE
linkToHtml http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwlV3da9RAEB9KLagPtraKp7VuwRfB9ZLd5Dbrm9jWBtuj2DvoW9jsBx6cSWkuov-9s3tJqCiFvoVksiw7OzO_2dmZAXjrtLCJQu3nVMppIpmgpRUTqoXivESBMtqFZhNiOs2uruTFBrwfcmGsteHymf3gH0Ms39S69UdlY0TbCBdQ4T5Ik4RF62ytIWbgW86G2GYqqEDg3wclIzme5RfoCrIYPVSGBta3q-FJqAU3-csehQYr_2jlYGpOtu83yR140kFK8mm9B57Chq12YbuDl6QT3mYXHt-qPbgH348u6Tz_SI5ataSX7bXXGQ3Sny9--aACqR35otrG51gO73zjtGVDEOeSOY4ZLhOsfpO8Txoki4rkP1BDkW_9vaS6egbzk-PZ51PatV2gGtHViiKL0tJxw2NdameljowpeaaZVYlzxiSJy5jTRkjGrFDMaGl8dE5xBH8a8c9z2Kzqyr4AIo0zIuPcupjjjkilsqaUEaK0eJJFko1g3C9-obua5L41xrIIvkkkC-Rc4TlXdJwbwbvhj-t1PY47aPc8ewa6jjMj2O_5W3QC2xToBaZoprmIRnA4fEZR8_ETVdm6DTRoveMJy17-f-Q38PB0dn5WnOXTr6_gkZ_L-rRmHzZXN619DVv652rR3ByEnfsHn5vqGg
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=article&rft.atitle=DS-UI%3A+Dual-Supervised+Mixture+of+Gaussian+Mixture+Models+for+Uncertainty+Inference+in+Image+Recognition&rft.jtitle=IEEE+transactions+on+image+processing&rft.au=Xie%2C+Jiyang&rft.au=Ma%2C+Zhanyu&rft.au=Xue%2C+Jing-Hao&rft.au=Zhang%2C+Guoqiang&rft.date=2021&rft.pub=IEEE&rft.issn=1057-7149&rft.volume=30&rft.spage=9208&rft.epage=9219&rft_id=info:doi/10.1109%2FTIP.2021.3123555&rft_id=info%3Apmid%2F34739376&rft.externalDocID=9605222
thumbnail_l http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/lc.gif&issn=1057-7149&client=summon
thumbnail_m http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/mc.gif&issn=1057-7149&client=summon
thumbnail_s http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/sc.gif&issn=1057-7149&client=summon