Aggregated Contextual Transformations for High-Resolution Image Inpainting

Image inpainting that completes large free-form missing regions in images is a promising yet challenging task. State-of-the-art approaches have achieved significant progress by taking advantage of generative adversarial networks (GAN). However, these approaches can suffer from generating distorted s...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:IEEE transactions on visualization and computer graphics Jg. 29; H. 7; S. 3266 - 3280
Hauptverfasser: Zeng, Yanhong, Fu, Jianlong, Chao, Hongyang, Guo, Baining
Format: Journal Article
Sprache:Englisch
Veröffentlicht: United States IEEE 01.07.2023
The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
Schlagworte:
Cognition
Context
Convolution
Discriminators
Filling
Free form
Generative adversarial networks
generative adversarial networks (GAN)
Generators
High resolution
Image enhancement
image inpainting
Image resolution
Image synthesis
object removal
Reasoning
State-of-the-art reviews
Synthesis
Task analysis
Texture
Training
ISSN:1077-2626, 1941-0506, 1941-0506
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Abstract Image inpainting that completes large free-form missing regions in images is a promising yet challenging task. State-of-the-art approaches have achieved significant progress by taking advantage of generative adversarial networks (GAN). However, these approaches can suffer from generating distorted structures and blurry textures in high-resolution images (e.g., <inline-formula><tex-math notation="LaTeX">512\times 512</tex-math> <mml:math><mml:mrow><mml:mn>512</mml:mn><mml:mo>×</mml:mo><mml:mn>512</mml:mn></mml:mrow></mml:math><inline-graphic xlink:href="fu-ieq1-3156949.gif"/> </inline-formula>). The challenges mainly drive from (1) image content reasoning from distant contexts, and (2) fine-grained texture synthesis for a large missing region. To overcome these two challenges, we propose an enhanced GAN-based model, named A ggregated C O ntextual- T ransformation GAN ( AOT-GAN ), for high-resolution image inpainting. Specifically, to enhance context reasoning, we construct the generator of AOT-GAN by stacking multiple layers of a proposed AOT block. The AOT blocks aggregate contextual transformations from various receptive fields, allowing to capture both informative distant image contexts and rich patterns of interest for context reasoning. For improving texture synthesis, we enhance the discriminator of AOT-GAN by training it with a tailored mask-prediction task. Such a training objective forces the discriminator to distinguish the detailed appearances of real and synthesized patches, and in turn facilitates the generator to synthesize clear textures. Extensive comparisons on Places2, the most challenging benchmark with 1.8 million high-resolution images of 365 complex scenes, show that our model outperforms the state-of-the-art. A user study including more than 30 subjects further validates the superiority of AOT-GAN. We further evaluate the proposed AOT-GAN in practical applications, e.g., logo removal, face editing, and object removal. Results show that our model achieves promising completions in the real world. We release codes and models in https://github.com/researchmm/AOT-GAN-for-Inpainting .
AbstractList Image inpainting that completes large free-form missing regions in images is a promising yet challenging task. State-of-the-art approaches have achieved significant progress by taking advantage of generative adversarial networks (GAN). However, these approaches can suffer from generating distorted structures and blurry textures in high-resolution images (e.g., 512×512). The challenges mainly drive from (1) image content reasoning from distant contexts, and (2) fine-grained texture synthesis for a large missing region. To overcome these two challenges, we propose an enhanced GAN-based model, named Aggregated COntextual-Transformation GAN (AOT-GAN), for high-resolution image inpainting. Specifically, to enhance context reasoning, we construct the generator of AOT-GAN by stacking multiple layers of a proposed AOT block. The AOT blocks aggregate contextual transformations from various receptive fields, allowing to capture both informative distant image contexts and rich patterns of interest for context reasoning. For improving texture synthesis, we enhance the discriminator of AOT-GAN by training it with a tailored mask-prediction task. Such a training objective forces the discriminator to distinguish the detailed appearances of real and synthesized patches, and in turn facilitates the generator to synthesize clear textures. Extensive comparisons on Places2, the most challenging benchmark with 1.8 million high-resolution images of 365 complex scenes, show that our model outperforms the state-of-the-art. A user study including more than 30 subjects further validates the superiority of AOT-GAN. We further evaluate the proposed AOT-GAN in practical applications, e.g., logo removal, face editing, and object removal. Results show that our model achieves promising completions in the real world. We release codes and models in https://github.com/researchmm/AOT-GAN-for-Inpainting.Image inpainting that completes large free-form missing regions in images is a promising yet challenging task. State-of-the-art approaches have achieved significant progress by taking advantage of generative adversarial networks (GAN). However, these approaches can suffer from generating distorted structures and blurry textures in high-resolution images (e.g., 512×512). The challenges mainly drive from (1) image content reasoning from distant contexts, and (2) fine-grained texture synthesis for a large missing region. To overcome these two challenges, we propose an enhanced GAN-based model, named Aggregated COntextual-Transformation GAN (AOT-GAN), for high-resolution image inpainting. Specifically, to enhance context reasoning, we construct the generator of AOT-GAN by stacking multiple layers of a proposed AOT block. The AOT blocks aggregate contextual transformations from various receptive fields, allowing to capture both informative distant image contexts and rich patterns of interest for context reasoning. For improving texture synthesis, we enhance the discriminator of AOT-GAN by training it with a tailored mask-prediction task. Such a training objective forces the discriminator to distinguish the detailed appearances of real and synthesized patches, and in turn facilitates the generator to synthesize clear textures. Extensive comparisons on Places2, the most challenging benchmark with 1.8 million high-resolution images of 365 complex scenes, show that our model outperforms the state-of-the-art. A user study including more than 30 subjects further validates the superiority of AOT-GAN. We further evaluate the proposed AOT-GAN in practical applications, e.g., logo removal, face editing, and object removal. Results show that our model achieves promising completions in the real world. We release codes and models in https://github.com/researchmm/AOT-GAN-for-Inpainting.
Image inpainting that completes large free-form missing regions in images is a promising yet challenging task. State-of-the-art approaches have achieved significant progress by taking advantage of generative adversarial networks (GAN). However, these approaches can suffer from generating distorted structures and blurry textures in high-resolution images (e.g., <inline-formula><tex-math notation="LaTeX">512\times 512</tex-math> <mml:math><mml:mrow><mml:mn>512</mml:mn><mml:mo>×</mml:mo><mml:mn>512</mml:mn></mml:mrow></mml:math><inline-graphic xlink:href="fu-ieq1-3156949.gif"/> </inline-formula>). The challenges mainly drive from (1) image content reasoning from distant contexts, and (2) fine-grained texture synthesis for a large missing region. To overcome these two challenges, we propose an enhanced GAN-based model, named A ggregated C O ntextual- T ransformation GAN ( AOT-GAN ), for high-resolution image inpainting. Specifically, to enhance context reasoning, we construct the generator of AOT-GAN by stacking multiple layers of a proposed AOT block. The AOT blocks aggregate contextual transformations from various receptive fields, allowing to capture both informative distant image contexts and rich patterns of interest for context reasoning. For improving texture synthesis, we enhance the discriminator of AOT-GAN by training it with a tailored mask-prediction task. Such a training objective forces the discriminator to distinguish the detailed appearances of real and synthesized patches, and in turn facilitates the generator to synthesize clear textures. Extensive comparisons on Places2, the most challenging benchmark with 1.8 million high-resolution images of 365 complex scenes, show that our model outperforms the state-of-the-art. A user study including more than 30 subjects further validates the superiority of AOT-GAN. We further evaluate the proposed AOT-GAN in practical applications, e.g., logo removal, face editing, and object removal. Results show that our model achieves promising completions in the real world. We release codes and models in https://github.com/researchmm/AOT-GAN-for-Inpainting .
Image inpainting that completes large free-form missing regions in images is a promising yet challenging task. State-of-the-art approaches have achieved significant progress by taking advantage of generative adversarial networks (GAN). However, these approaches can suffer from generating distorted structures and blurry textures in high-resolution images (e.g., [Formula Omitted]). The challenges mainly drive from (1) image content reasoning from distant contexts, and (2) fine-grained texture synthesis for a large missing region. To overcome these two challenges, we propose an enhanced GAN-based model, named A ggregated C O ntextual- T ransformation GAN ( AOT-GAN ), for high-resolution image inpainting. Specifically, to enhance context reasoning, we construct the generator of AOT-GAN by stacking multiple layers of a proposed AOT block. The AOT blocks aggregate contextual transformations from various receptive fields, allowing to capture both informative distant image contexts and rich patterns of interest for context reasoning. For improving texture synthesis, we enhance the discriminator of AOT-GAN by training it with a tailored mask-prediction task. Such a training objective forces the discriminator to distinguish the detailed appearances of real and synthesized patches, and in turn facilitates the generator to synthesize clear textures. Extensive comparisons on Places2, the most challenging benchmark with 1.8 million high-resolution images of 365 complex scenes, show that our model outperforms the state-of-the-art. A user study including more than 30 subjects further validates the superiority of AOT-GAN. We further evaluate the proposed AOT-GAN in practical applications, e.g., logo removal, face editing, and object removal. Results show that our model achieves promising completions in the real world. We release codes and models in https://github.com/researchmm/AOT-GAN-for-Inpainting .
Image inpainting that completes large free-form missing regions in images is a promising yet challenging task. State-of-the-art approaches have achieved significant progress by taking advantage of generative adversarial networks (GAN). However, these approaches can suffer from generating distorted structures and blurry textures in high-resolution images (e.g., 512×512). The challenges mainly drive from (1) image content reasoning from distant contexts, and (2) fine-grained texture synthesis for a large missing region. To overcome these two challenges, we propose an enhanced GAN-based model, named Aggregated COntextual-Transformation GAN (AOT-GAN), for high-resolution image inpainting. Specifically, to enhance context reasoning, we construct the generator of AOT-GAN by stacking multiple layers of a proposed AOT block. The AOT blocks aggregate contextual transformations from various receptive fields, allowing to capture both informative distant image contexts and rich patterns of interest for context reasoning. For improving texture synthesis, we enhance the discriminator of AOT-GAN by training it with a tailored mask-prediction task. Such a training objective forces the discriminator to distinguish the detailed appearances of real and synthesized patches, and in turn facilitates the generator to synthesize clear textures. Extensive comparisons on Places2, the most challenging benchmark with 1.8 million high-resolution images of 365 complex scenes, show that our model outperforms the state-of-the-art. A user study including more than 30 subjects further validates the superiority of AOT-GAN. We further evaluate the proposed AOT-GAN in practical applications, e.g., logo removal, face editing, and object removal. Results show that our model achieves promising completions in the real world. We release codes and models in https://github.com/researchmm/AOT-GAN-for-Inpainting.
Author Fu, Jianlong
Chao, Hongyang
Zeng, Yanhong
Guo, Baining
Author_xml – sequence: 1
  givenname: Yanhong
  orcidid: 0000-0003-3596-5163
  surname: Zeng
  fullname: Zeng, Yanhong
  email: zengyh7@mail2.sysu.edu.cn
  organization: School of Computer Science and Engineering, Sun Yat-sen University, Guangzhou, China
– sequence: 2
  givenname: Jianlong
  orcidid: 0000-0002-1025-2012
  surname: Fu
  fullname: Fu, Jianlong
  email: jianf@microsoft.com
  organization: Microsoft Research, Redmond, WA, USA
– sequence: 3
  givenname: Hongyang
  orcidid: 0000-0002-6104-2322
  surname: Chao
  fullname: Chao, Hongyang
  email: isschhy@mail.sysu.edu.cn
  organization: School of Computer Science and Engineering, Sun Yat-sen University, Guangzhou, China
– sequence: 4
  givenname: Baining
  surname: Guo
  fullname: Guo, Baining
  email: bainguo@microsoft.com
  organization: Microsoft Research, Redmond, WA, USA
BackLink https://www.ncbi.nlm.nih.gov/pubmed/35254985$$D View this record in MEDLINE/PubMed
BookMark eNp9kU1rGzEQhkVwSZyPHxAKZSGXXNbV966OxqSJSyAQ3FyFLM9uFXYlV9JC---7jp0cfMhphuF5Z4b3PUcTHzwgdE3wjBCsvq9eFvcziimdMSKk4uoETYnipMQCy8nY46oqqaTyDJ2n9Iox4bxWp-iMCSq4qsUU_Zy3bYTWZNgUi-Az_M2D6YpVND41IfYmu-BTMbbFg2t_l8-QQjfshsWyNy0US781zmfn20v0pTFdgqtDvUC_ftytFg_l49P9cjF_LC3jKpd1gymuLQfGBQMl11JUDbXUYrsm1ljRUKAbRvGGMyzAVrZq1kQ21tYYC2LYBbrd793G8GeAlHXvkoWuMx7CkDSVrGJEKaFG9OYIfQ1D9ON3mtZ0tIdwJUbq24Ea1j1s9Da63sR_-t2lESB7wMaQUoTmAyFY75LQuyT0Lgl9SGLUVEca6_Kbmzka132q_LpXOgD4uKQqqoTk7D8xe5T3
CODEN ITVGEA
CitedBy_id crossref_primary_10_3390_electronics12061490
crossref_primary_10_1016_j_icte_2024_08_008
crossref_primary_10_1016_j_asoc_2025_113271
crossref_primary_10_1109_LSENS_2024_3404139
crossref_primary_10_1038_s40494_025_01693_z
crossref_primary_10_1016_j_image_2025_117326
crossref_primary_10_1117_1_JEI_33_5_053011
crossref_primary_10_1038_s40494_025_01710_1
crossref_primary_10_1109_JSTARS_2025_3547917
crossref_primary_10_3390_rs15092454
crossref_primary_10_3390_app132011189
crossref_primary_10_1109_ACCESS_2024_3443924
crossref_primary_10_3390_electronics13101852
crossref_primary_10_3390_s24144470
crossref_primary_10_1109_JETCAS_2025_3575272
crossref_primary_10_1109_TMM_2024_3521800
crossref_primary_10_1038_s40494_025_01891_9
crossref_primary_10_1109_JSEN_2022_3211007
crossref_primary_10_3389_frai_2025_1614608
crossref_primary_10_3390_jmse12122178
crossref_primary_10_1109_ACCESS_2024_3484472
crossref_primary_10_1109_TGRS_2024_3390392
crossref_primary_10_1109_JBHI_2023_3288830
crossref_primary_10_1088_1361_6560_aced78
crossref_primary_10_1109_TCSVT_2025_3532321
crossref_primary_10_3390_ani15070978
crossref_primary_10_1007_s11220_025_00624_3
crossref_primary_10_1109_ACCESS_2025_3595120
crossref_primary_10_3390_math11244908
crossref_primary_10_1007_s00371_024_03549_2
crossref_primary_10_1038_s41598_024_82097_0
crossref_primary_10_3390_rs17132132
crossref_primary_10_1049_ipr2_13318
crossref_primary_10_3390_math11204349
crossref_primary_10_1016_j_displa_2025_103092
crossref_primary_10_1016_j_neucom_2025_131001
crossref_primary_10_1016_j_ijmecsci_2025_110797
crossref_primary_10_1186_s13640_025_00670_7
crossref_primary_10_1007_s10462_024_11075_9
crossref_primary_10_1109_TVCG_2023_3286394
crossref_primary_10_1016_j_displa_2024_102945
crossref_primary_10_1360_SST_2024_0343
crossref_primary_10_3389_fphy_2023_1277710
crossref_primary_10_1016_j_ymssp_2024_111783
crossref_primary_10_2478_ijanmc_2024_0031
crossref_primary_10_3389_frai_2025_1616007
crossref_primary_10_1007_s11042_024_19365_8
crossref_primary_10_1007_s00530_024_01585_5
crossref_primary_10_3390_math13142254
crossref_primary_10_3390_math11173644
crossref_primary_10_1007_s11760_023_02931_2
crossref_primary_10_1007_s11042_025_21027_2
crossref_primary_10_1007_s00371_023_03045_z
crossref_primary_10_1038_s41598_024_51651_1
crossref_primary_10_1007_s00371_025_04170_7
crossref_primary_10_1016_j_eswa_2025_128957
crossref_primary_10_3390_jmse13050997
crossref_primary_10_1016_j_neucom_2025_130133
crossref_primary_10_1109_TCSVT_2024_3370578
crossref_primary_10_3390_s23239439
crossref_primary_10_1007_s11760_025_03987_y
crossref_primary_10_1016_j_engappai_2024_109397
crossref_primary_10_1007_s00530_024_01466_x
crossref_primary_10_1038_s41377_023_01296_y
crossref_primary_10_1016_j_neucom_2025_129663
crossref_primary_10_1109_ACCESS_2023_3301568
crossref_primary_10_1007_s11760_023_02939_8
crossref_primary_10_1038_s41598_024_72368_1
crossref_primary_10_1109_ACCESS_2023_3285798
crossref_primary_10_1016_j_bspc_2025_108296
crossref_primary_10_1007_s11760_024_03415_7
crossref_primary_10_1109_TVCG_2025_3550844
crossref_primary_10_3390_info14090512
crossref_primary_10_1007_s00530_024_01609_0
crossref_primary_10_1016_j_bdes_2025_100004
crossref_primary_10_32604_cmc_2022_023071
crossref_primary_10_1016_j_cviu_2025_104495
crossref_primary_10_3390_app13137638
crossref_primary_10_1007_s11263_025_02448_w
crossref_primary_10_1038_s40494_025_01719_6
crossref_primary_10_1049_cmu2_12846
crossref_primary_10_1016_j_engappai_2024_109181
crossref_primary_10_1109_TCSVT_2024_3454457
crossref_primary_10_1088_1361_6560_ad02d9
crossref_primary_10_1007_s11760_025_04685_5
crossref_primary_10_1049_ipr2_70105
crossref_primary_10_3390_electronics13101873
crossref_primary_10_1364_OL_571252
crossref_primary_10_1063_5_0139034
crossref_primary_10_1109_TPAMI_2025_3558092
crossref_primary_10_1007_s00371_024_03578_x
crossref_primary_10_3233_JIFS_239513
crossref_primary_10_1007_s00371_024_03702_x
crossref_primary_10_1109_JSTARS_2025_3591103
crossref_primary_10_1038_s40494_025_01607_z
crossref_primary_10_1109_LSP_2023_3340998
crossref_primary_10_1109_TWC_2024_3509382
crossref_primary_10_1016_j_eswa_2025_129667
crossref_primary_10_1049_ipr2_12958
crossref_primary_10_1109_TCSVT_2024_3450493
Cites_doi 10.1109/ICCV48922.2021.00465
10.1109/TPAMI.2007.60
10.1109/ICCV.2017.89
10.1007/978-3-030-01252-6_6
10.1109/CVPR.2016.90
10.1109/TVCG.2014.2298016
10.1007/978-3-030-66823-5_1
10.1109/ICCV.2019.00457
10.1109/CVPR.2016.278
10.1109/CVPR42600.2020.00583
10.1109/ACSSC.2003.1292216
10.1109/ICCV48922.2021.01008
10.1109/CVPR.2017.624
10.1109/CVPR.2017.434
10.1109/CVPR.2001.990497
10.1109/ICCV48922.2021.01434
10.1145/3072959.3073659
10.1109/ICCV.2001.937493
10.1109/TVCG.2015.2462368
10.1109/TIP.2004.833105
10.1109/CVPR.2009.5206848
10.1109/TPAMI.2017.2723009
10.1109/CVPR.2017.634
10.1109/TVCG.2018.2889297
10.1109/CVPR.2017.728
10.1109/ICCV.2019.00027
10.1109/CVPR.2018.00577
10.1109/ICCV.2017.304
10.1109/TVCG.2017.2702738
10.1109/ICCV.2019.00895
10.1109/CVPR.2019.01162
10.1145/501786.501787
10.1109/CVPR.2019.00158
10.1145/1073204.1073274
10.1109/CVPR.2019.00953
10.1145/2980179.2982398
10.1145/344779.344972
10.1007/978-3-030-58517-4_31
10.1109/TIP.2003.815261
10.1016/j.image.2017.10.003
10.1145/3474085.3475421
10.1109/WACV.2018.00163
10.1109/CVPR.2019.00599
10.1007/978-3-030-01216-8_1
10.1007/s00530-005-0167-6
10.1007/978-3-030-01264-9_1
10.1109/CVPR.2017.632
10.1109/CVPR.2018.00068
10.1016/j.cviu.2020.103155
10.1109/CVPR42600.2020.00753
10.1109/CVPR.2016.265
10.1145/1531326.1531330
10.1145/3394171.3413853
10.1109/ICCV.1999.790383
10.1109/TPAMI.2017.2699184
10.1145/3219819.3219944
10.1145/383259.383296
10.1080/00031305.1994.10476030
10.1109/ICCV.2019.00427
ContentType Journal Article
Copyright Copyright The Institute of Electrical and Electronics Engineers, Inc. (IEEE) 2023
Copyright_xml – notice: Copyright The Institute of Electrical and Electronics Engineers, Inc. (IEEE) 2023
DBID 97E
RIA
RIE
AAYXX
CITATION
NPM
7SC
7SP
8FD
JQ2
L7M
L~C
L~D
7X8
DOI 10.1109/TVCG.2022.3156949
DatabaseName IEEE All-Society Periodicals Package (ASPP) 2005-present
IEEE All-Society Periodicals Package (ASPP) 1998-Present
IEEE Electronic Library (IEL)
CrossRef
PubMed
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
PubMed
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
PubMed
Database_xml – sequence: 1
  dbid: NPM
  name: PubMed
  url: http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed
  sourceTypes: Index Database
– sequence: 2
  dbid: RIE
  name: IEEE Electronic Library (IEL)
  url: https://ieeexplore.ieee.org/
  sourceTypes: Publisher
– sequence: 3
  dbid: 7X8
  name: MEDLINE - Academic
  url: https://search.proquest.com/medline
  sourceTypes: Aggregation Database
DeliveryMethod fulltext_linktorsrc
Discipline Engineering
EISSN 1941-0506
EndPage 3280
ExternalDocumentID 35254985
10_1109_TVCG_2022_3156949
9729564
Genre orig-research
Journal Article
GrantInformation_xml – fundername: National Natural Science Foundation of China; NSF of China
  grantid: 61672548; U1611461
  funderid: 10.13039/501100001809
GroupedDBID ---
-~X
.DC
0R~
29I
4.4
53G
5GY
6IK
97E
AAJGR
AARMG
AASAJ
AAWTH
ABAZT
ABQJQ
ABVLG
ACGFO
ACIWK
AENEX
AGQYO
AHBIQ
AKJIK
AKQYR
ALMA_UNASSIGNED_HOLDINGS
ATWAV
BEFXN
BFFAM
BGNUA
BKEBE
BPEOZ
CS3
DU5
EBS
EJD
F5P
HZ~
IEDLZ
IFIPE
IPLJI
JAVBF
LAI
M43
O9-
OCL
P2P
PQQKQ
RIA
RIE
RNS
TN5
AAYXX
CITATION
5VS
AETIX
AGSQL
AI.
AIBXA
ALLEH
H~9
IFJZH
NPM
RIG
RNI
RZB
VH1
7SC
7SP
8FD
JQ2
L7M
L~C
L~D
7X8
ID FETCH-LOGICAL-c349t-8f0208c4e3453e96b657f2c2c0cb1cac5f2e2d320d4305ec7c7fb16fcc80051a3
IEDL.DBID RIE
ISICitedReferencesCount 162
ISICitedReferencesURI http://www.webofscience.com/api/gateway?GWVersion=2&SrcApp=Summon&SrcAuth=ProQuest&DestLinkType=CitingArticles&DestApp=WOS_CPL&KeyUT=001000210200001&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
ISSN 1077-2626
1941-0506
IngestDate Thu Oct 02 03:26:54 EDT 2025
Sun Jun 29 12:32:10 EDT 2025
Mon Jul 21 06:03:37 EDT 2025
Sat Nov 29 03:31:40 EST 2025
Tue Nov 18 22:22:54 EST 2025
Wed Aug 27 02:14:17 EDT 2025
IsPeerReviewed true
IsScholarly true
Issue 7
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-c349t-8f0208c4e3453e96b657f2c2c0cb1cac5f2e2d320d4305ec7c7fb16fcc80051a3
Notes ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
content type line 23
ORCID 0000-0002-1025-2012
0000-0002-6104-2322
0000-0003-3596-5163
PMID 35254985
PQID 2821071495
PQPubID 75741
PageCount 15
ParticipantIDs proquest_miscellaneous_2637319959
ieee_primary_9729564
crossref_citationtrail_10_1109_TVCG_2022_3156949
pubmed_primary_35254985
proquest_journals_2821071495
crossref_primary_10_1109_TVCG_2022_3156949
PublicationCentury 2000
PublicationDate 2023-07-01
PublicationDateYYYYMMDD 2023-07-01
PublicationDate_xml – month: 07
  year: 2023
  text: 2023-07-01
  day: 01
PublicationDecade 2020
PublicationPlace United States
PublicationPlace_xml – name: United States
– name: New York
PublicationTitle IEEE transactions on visualization and computer graphics
PublicationTitleAbbrev TVCG
PublicationTitleAlternate IEEE Trans Vis Comput Graph
PublicationYear 2023
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 ref13
miyato (ref52) 2018
zeng (ref37) 2021
ref12
ref56
ref15
ref59
pukelsheim (ref67) 1994; 48
ref14
ref58
ref53
ref11
ref55
ref10
ref54
ref19
ref18
simonyan (ref57) 2014
ref51
ref50
ref46
ref48
ref47
ref42
ref41
ref44
ref43
yu (ref45) 2016
ref49
ref8
ref7
ref9
ref4
ref3
ref6
ref5
nazeri (ref17) 2019
ref40
ref35
heusel (ref65) 2017
ref34
ref36
ref31
ref30
ref33
ref32
goodfellow (ref16) 2014
ref2
ref1
zhao (ref60) 2020
ref39
wang (ref20) 2018
karras (ref61) 2018
ref38
su (ref62) 2018
johnson (ref27) 2016
ref71
ref70
ref72
ref24
ref68
ronneberger (ref63) 2015
ref23
ref26
ref25
ref69
ref64
ref22
ref66
ref21
ref28
ref29
References_xml – ident: ref59
  doi: 10.1109/ICCV48922.2021.00465
– start-page: 2672
  year: 2014
  ident: ref16
  article-title: Generative adversarial nets
  publication-title: Proc 27th Int Conf Neural Informat Process Syst
– ident: ref4
  doi: 10.1109/TPAMI.2007.60
– year: 2016
  ident: ref45
  article-title: Multi-scale context aggregation by dilated convolutions
  publication-title: Proc Int Conf Learn Representations
– ident: ref70
  doi: 10.1109/ICCV.2017.89
– year: 2021
  ident: ref37
  article-title: Improving visual quality of image synthesis by a token-based generator with transformers
  publication-title: Proc Int Conf Neural Informat Process Syst
– ident: ref50
  doi: 10.1007/978-3-030-01252-6_6
– start-page: 694
  year: 2016
  ident: ref27
  article-title: Perceptual losses for real-time style transfer and super-resolution
  publication-title: Proc Eur Conf Comput Vis
– ident: ref15
  doi: 10.1109/CVPR.2016.90
– ident: ref6
  doi: 10.1109/TVCG.2014.2298016
– ident: ref54
  doi: 10.1007/978-3-030-66823-5_1
– ident: ref7
  doi: 10.1109/ICCV.2019.00457
– ident: ref14
  doi: 10.1109/CVPR.2016.278
– ident: ref36
  doi: 10.1109/CVPR42600.2020.00583
– ident: ref64
  doi: 10.1109/ACSSC.2003.1292216
– ident: ref35
  doi: 10.1109/ICCV48922.2021.01008
– ident: ref38
  doi: 10.1109/CVPR.2017.624
– ident: ref19
  doi: 10.1109/CVPR.2017.434
– ident: ref30
  doi: 10.1109/CVPR.2001.990497
– ident: ref49
  doi: 10.1109/ICCV48922.2021.01434
– ident: ref40
  doi: 10.1145/3072959.3073659
– start-page: 6626
  year: 2017
  ident: ref65
  article-title: GANs trained by a two time-scale update rule converge to a local nash equilibrium
  publication-title: Proc 31st Int Conf Neural Informat Process Syst
– ident: ref29
  doi: 10.1109/ICCV.2001.937493
– ident: ref5
  doi: 10.1109/TVCG.2015.2462368
– ident: ref3
  doi: 10.1109/TIP.2004.833105
– ident: ref58
  doi: 10.1109/CVPR.2009.5206848
– ident: ref26
  doi: 10.1109/TPAMI.2017.2723009
– ident: ref53
  doi: 10.1109/CVPR.2017.634
– ident: ref10
  doi: 10.1109/TVCG.2018.2889297
– start-page: 234
  year: 2015
  ident: ref63
  article-title: U-net: Convolutional networks for biomedical image segmentation
  publication-title: Proc Int Conf Med Image Comput Comput -Assist Interv
– start-page: 331
  year: 2018
  ident: ref20
  article-title: Image inpainting via generative multi-column convolutional neural networks
  publication-title: Proc Int Conf Neural Informat Process Syst
– ident: ref39
  doi: 10.1109/CVPR.2017.728
– ident: ref51
  doi: 10.1109/ICCV.2019.00027
– year: 2018
  ident: ref52
  article-title: Spectral normalization for generative adversarial networks
  publication-title: Proc Int Conf Learn Representations
– ident: ref13
  doi: 10.1109/CVPR.2018.00577
– ident: ref56
  doi: 10.1109/ICCV.2017.304
– ident: ref11
  doi: 10.1109/TVCG.2017.2702738
– ident: ref42
  doi: 10.1109/ICCV.2019.00895
– ident: ref44
  doi: 10.1109/CVPR.2019.01162
– year: 2018
  ident: ref62
  article-title: Open logo detection challenge
  publication-title: Proc Brit Mach Vis Conf
– ident: ref31
  doi: 10.1145/501786.501787
– ident: ref21
  doi: 10.1109/CVPR.2019.00158
– ident: ref12
  doi: 10.1145/1073204.1073274
– year: 2019
  ident: ref17
  article-title: EdgeConnect: Generative image inpainting with adversarial edge learning
  publication-title: Proc Int Conf Comput Vis
– ident: ref71
  doi: 10.1109/CVPR.2019.00953
– ident: ref72
  doi: 10.1145/2980179.2982398
– ident: ref1
  doi: 10.1145/344779.344972
– ident: ref18
  doi: 10.1007/978-3-030-58517-4_31
– ident: ref9
  doi: 10.1109/TIP.2003.815261
– ident: ref69
  doi: 10.1016/j.image.2017.10.003
– ident: ref34
  doi: 10.1145/3474085.3475421
– ident: ref23
  doi: 10.1109/WACV.2018.00163
– ident: ref25
  doi: 10.1109/CVPR.2019.00599
– ident: ref43
  doi: 10.1007/978-3-030-01216-8_1
– ident: ref68
  doi: 10.1007/s00530-005-0167-6
– ident: ref41
  doi: 10.1007/978-3-030-01264-9_1
– ident: ref24
  doi: 10.1109/CVPR.2017.632
– ident: ref66
  doi: 10.1109/CVPR.2018.00068
– ident: ref55
  doi: 10.1016/j.cviu.2020.103155
– ident: ref8
  doi: 10.1109/CVPR42600.2020.00753
– ident: ref28
  doi: 10.1109/CVPR.2016.265
– ident: ref2
  doi: 10.1145/1531326.1531330
– ident: ref48
  doi: 10.1145/3394171.3413853
– ident: ref32
  doi: 10.1109/ICCV.1999.790383
– ident: ref46
  doi: 10.1109/TPAMI.2017.2699184
– ident: ref47
  doi: 10.1145/3219819.3219944
– ident: ref33
  doi: 10.1145/383259.383296
– volume: 48
  start-page: 88
  year: 1994
  ident: ref67
  article-title: The three sigma rule
  publication-title: The Amer Statistician
  doi: 10.1080/00031305.1994.10476030
– year: 2018
  ident: ref61
  article-title: Progressive growing of GANs for improved quality, stability, and variation
  publication-title: Proc Int Conf Learn Representations
– ident: ref22
  doi: 10.1109/ICCV.2019.00427
– year: 2014
  ident: ref57
  article-title: Very deep convolutional networks for large-scale image recognition
– year: 2020
  ident: ref60
  article-title: Large scale image completion via co-modulated generative adversarial networks
  publication-title: Proc Int Conf Learn Representations
SSID ssj0014489
Score 2.694736
Snippet Image inpainting that completes large free-form missing regions in images is a promising yet challenging task. State-of-the-art approaches have achieved...
SourceID proquest
pubmed
crossref
ieee
SourceType Aggregation Database
Index Database
Enrichment Source
Publisher
StartPage 3266
SubjectTerms Cognition
Context
Convolution
Discriminators
Filling
Free form
Generative adversarial networks
generative adversarial networks (GAN)
Generators
High resolution
Image enhancement
image inpainting
Image resolution
Image synthesis
object removal
Reasoning
State-of-the-art reviews
Synthesis
Task analysis
Texture
Training
Title Aggregated Contextual Transformations for High-Resolution Image Inpainting
URI https://ieeexplore.ieee.org/document/9729564
https://www.ncbi.nlm.nih.gov/pubmed/35254985
https://www.proquest.com/docview/2821071495
https://www.proquest.com/docview/2637319959
Volume 29
WOSCitedRecordID wos001000210200001&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 Electronic Library (IEL)
  customDbUrl:
  eissn: 1941-0506
  dateEnd: 99991231
  omitProxy: false
  ssIdentifier: ssj0014489
  issn: 1077-2626
  databaseCode: RIE
  dateStart: 19950101
  isFulltext: true
  titleUrlDefault: https://ieeexplore.ieee.org/
  providerName: IEEE
link http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwlV1NT9wwEB3Bqgc40JaPspSiIPWEMCS2Y8fHFSotFUIctmhvUTJxEBLNIthF_HxmnGy0hxaJW6Q4TuIZ2zOemfcAvqNyxlUZCtTGCI1Gi6LUpaDluFRJpTEpAojrpb26yiYTd70Cx30tjPc-JJ_5E74MsfxqinM-Kjt1ZAmmRq_CqrW2rdXqIwbkZrg2v9AKSVZ6F8FMYnc6vjn7SZ6glOSgpsZpRgplFFDtmEF5aTsK_Cr_NzXDlnP-8X0f-wk2OtMyGrW68BlWfLMJ60uAg1vwe3RL_jWfnFVRwKV64fKRaLxkvZIWRnQZcQKI4MP9VjWji7-08kQXzUNxF8gltuHP-Y_x2S_RsSkIVNrNRFYzHydqr3SqvDOlSW0tUWKMZYIFprX0slIyrhgFzKNFW5eJqREznrmF2oFBM238LkRaYlZQp_SLhdZIa0CV0BZYx96lsY_tEOLFoObYQY0z48V9HlyO2OUskpxFknciGcJR_8hDi7PxVuMtHu--YTfUQ9hfSC7vZuJTTi5lwkVaLh3CYX-b5hAHRorGT-fUxiiruFadev7SSrzve6Eoe_9-51dYYwL6NoF3Hwazx7n_Bh_weXb39HhAijrJDoKivgIDb-Gw
linkProvider IEEE
linkToHtml http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwlV1NT9wwEB1RQCocgJYCC7RNpZ6qGhLbceIjQnws3a44bCtuUTJxKiTIIthF_HxmnGy0B4rELVIcJ_GM7RnPzHsA31FZY8sUBWpjhEajRV7oQtByXKio1BjlHsR1kAyH6dWVvVyAn10tjHPOJ5-5A770sfxyjFM-Kju0ZAnGRr-DpVhrGTXVWl3MgBwN22QYJkKSnd7GMKPQHo7-Hp-RLygluaixsZqxQhkHVFvmUJ7bkDzDyv-NTb_pnK6_7XM3YK01LoOjRhs-wIKrP8LqHOTgJlwc_SMPm8_OysAjUz1xAUkwmrNfSQ8Dugw4BUTw8X6jnEH_ltaeoF_f5deeXuIT_Dk9GR2fi5ZPQaDSdiLSihk5UTulY-WsKUycVBIlhlhEmGNcSSdLJcOSccAcJphURWQqxJTnbq62YLEe124HAi0xzalT-sVca6RVoIxoE6xCZ-PQhUkPwtmgZtiCjTPnxU3mnY7QZiySjEWStSLpwY_ukbsGaeO1xps83l3Ddqh7sD-TXNbOxYeMnMqIy7Rs3INv3W2aRRwayWs3nlIboxLF1erU83Yj8a7vmaLsvvzOr_D-fPR7kA36w197sMJ09E067z4sTu6n7jMs4-Pk-uH-i1fXZ5dN5A8
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=Aggregated+Contextual+Transformations+for+High-Resolution+Image+Inpainting&rft.jtitle=IEEE+transactions+on+visualization+and+computer+graphics&rft.au=Zeng%2C+Yanhong&rft.au=Fu%2C+Jianlong&rft.au=Chao%2C+Hongyang&rft.au=Guo%2C+Baining&rft.date=2023-07-01&rft.issn=1077-2626&rft.eissn=1941-0506&rft.volume=29&rft.issue=7&rft.spage=3266&rft.epage=3280&rft_id=info:doi/10.1109%2FTVCG.2022.3156949&rft.externalDBID=n%2Fa&rft.externalDocID=10_1109_TVCG_2022_3156949
thumbnail_l http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/lc.gif&issn=1077-2626&client=summon
thumbnail_m http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/mc.gif&issn=1077-2626&client=summon
thumbnail_s http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/sc.gif&issn=1077-2626&client=summon