STD-Explain: Generalizing explanations for spatio-temporal graph convolutional networks based on spatio-temporal decoupled perturbation

Spatio-temporal graph convolutional networks utilize an alternating combination of one-dimensional ordinary convolution and graph convolution to extract spatio-temporal features. This alternation intertwines temporal and spatial features closely, leading to a tight coupling between them. The presenc...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Neurocomputing (Amsterdam) Jg. 657; S. 131539
Hauptverfasser: Li, Yanshan, Shi, Ting, He, Suixuan, Chen, Zhiyuan, Zhang, Li, Yu, Rui, Xie, Weixin
Format: Journal Article
Sprache:Englisch
Veröffentlicht: Elsevier B.V 07.12.2025
Schlagworte:
ISSN:0925-2312
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Abstract Spatio-temporal graph convolutional networks utilize an alternating combination of one-dimensional ordinary convolution and graph convolution to extract spatio-temporal features. This alternation intertwines temporal and spatial features closely, leading to a tight coupling between them. The presence of spatio-temporal coupling complicates the analysis of spatio-temporal data, posing challenges for existing explainability algorithms to effectively separate and interpret these intertwined features. Therefore, we propose STD-Explain, an explainable algorithm based on spatio-temporal decoupled perturbation, which employs a two-stage perturbation approach considering subgraph and node-level explanations. Firstly, targeting the spatio-temporal coupling issue in spatio-temporal graph convolutional networks, the algorithm proposes a temporal perturbation algorithm based on Slice Graph and a spatial perturbation algorithm aimed at important subgraph node features. Secondly, to avoid introducing additional semantic information when extracting temporal subgraphs, we propose a method for generating temporal subgraphs in spatio-temporal decoupling, slicing human skeleton sequences with discrete masks to ensure each subsequence maintains spatial structure integrity without introducing additional edges. Furthermore, to ensure the maximum correlation between the interpreted subgraphs and model predictions, we propose a temporal important subgraph discrimination strategy to select the most relevant subgraphs to model predictions. Experimental results demonstrate that STD-Explain performs well in qualitative and quantitative analysis.
AbstractList Spatio-temporal graph convolutional networks utilize an alternating combination of one-dimensional ordinary convolution and graph convolution to extract spatio-temporal features. This alternation intertwines temporal and spatial features closely, leading to a tight coupling between them. The presence of spatio-temporal coupling complicates the analysis of spatio-temporal data, posing challenges for existing explainability algorithms to effectively separate and interpret these intertwined features. Therefore, we propose STD-Explain, an explainable algorithm based on spatio-temporal decoupled perturbation, which employs a two-stage perturbation approach considering subgraph and node-level explanations. Firstly, targeting the spatio-temporal coupling issue in spatio-temporal graph convolutional networks, the algorithm proposes a temporal perturbation algorithm based on Slice Graph and a spatial perturbation algorithm aimed at important subgraph node features. Secondly, to avoid introducing additional semantic information when extracting temporal subgraphs, we propose a method for generating temporal subgraphs in spatio-temporal decoupling, slicing human skeleton sequences with discrete masks to ensure each subsequence maintains spatial structure integrity without introducing additional edges. Furthermore, to ensure the maximum correlation between the interpreted subgraphs and model predictions, we propose a temporal important subgraph discrimination strategy to select the most relevant subgraphs to model predictions. Experimental results demonstrate that STD-Explain performs well in qualitative and quantitative analysis.
ArticleNumber 131539
Author He, Suixuan
Chen, Zhiyuan
Xie, Weixin
Li, Yanshan
Zhang, Li
Yu, Rui
Shi, Ting
Author_xml – sequence: 1
  givenname: Yanshan
  orcidid: 0000-0002-8814-4628
  surname: Li
  fullname: Li, Yanshan
  email: lys@szu.edu.cn
  organization: Institute of Intelligent Information Processing, Shenzhen University, Shenzhen, 518060, China
– sequence: 2
  givenname: Ting
  surname: Shi
  fullname: Shi, Ting
  organization: Institute of Intelligent Information Processing, Shenzhen University, Shenzhen, 518060, China
– sequence: 3
  givenname: Suixuan
  surname: He
  fullname: He, Suixuan
  organization: Institute of Intelligent Information Processing, Shenzhen University, Shenzhen, 518060, China
– sequence: 4
  givenname: Zhiyuan
  surname: Chen
  fullname: Chen, Zhiyuan
  organization: Institute of Intelligent Information Processing, Shenzhen University, Shenzhen, 518060, China
– sequence: 5
  givenname: Li
  surname: Zhang
  fullname: Zhang, Li
  organization: Institute of Intelligent Information Processing, Shenzhen University, Shenzhen, 518060, China
– sequence: 6
  givenname: Rui
  surname: Yu
  fullname: Yu, Rui
  organization: Institute of Intelligent Information Processing, Shenzhen University, Shenzhen, 518060, China
– sequence: 7
  givenname: Weixin
  surname: Xie
  fullname: Xie, Weixin
  organization: Institute of Intelligent Information Processing, Shenzhen University, Shenzhen, 518060, China
BookMark eNp9UMtOwzAQ9KFItIU_4OAfSLCdV8MBCZVSkCpxoJwtPzbFJbUjOymPH-C3SQg3JPay2tmd2d2ZoYl1FhC6oCSmhOaX-9hCp9whZoRlMU1olpQTNCUlyyKWUHaKZiHsCaEFZeUUfT1tb6PVe1MLY6_wGix4UZtPY3cYBtSK1jgbcOU8Ds1QRC0cGtdP4Z0XzQtWzh5d3Q1jPWahfXP-NWApAmjs7B-WBuW6pu6bDfi28_Jnwxk6qUQd4Pw3z9Hz3Wq7vI82j-uH5c0mUiwr2kj0L-aq0KClVomkBeujgqKEJKcZW6RapEprIheaQb4gZVYJBYlMU81kKUUyR-moq7wLwUPFG28Own9wSvhgIN_z0UA-GMhHA3va9UiD_rajAc-DMmAVaONBtVw787_AN68DhLw
Cites_doi 10.1109/TNNLS.2021.3061115
10.1016/j.inffus.2025.103387
10.1109/TMM.2021.3121559
10.1109/TMM.2021.3086758
10.1007/s11263-025-02393-8
10.1109/TPAMI.2021.3115452
10.1109/TPAMI.2019.2916873
10.1109/TKDE.2022.3187455
10.1109/TMM.2022.3233442
10.1016/j.patcog.2020.107293
10.1109/TAFFC.2022.3181053
10.1109/TPAMI.2024.3413026
10.1109/TPAMI.2024.3367416
10.1016/j.patcog.2024.110251
10.1109/TNNLS.2022.3152990
10.1007/s10462-020-09904-8
10.1109/TKDE.2022.3201170
10.1016/j.neucom.2024.128393
ContentType Journal Article
Copyright 2025
Copyright_xml – notice: 2025
DBID AAYXX
CITATION
DOI 10.1016/j.neucom.2025.131539
DatabaseName CrossRef
DatabaseTitle CrossRef
DatabaseTitleList
DeliveryMethod fulltext_linktorsrc
Discipline Computer Science
ExternalDocumentID 10_1016_j_neucom_2025_131539
S0925231225022118
GrantInformation_xml – fundername: Department of Education of Guangdong Province
  grantid: 2020KCXTD004
  funderid: https://doi.org/10.13039/501100010226
– fundername: National Natural Science Foundation of China
  grantid: 62076165
  funderid: https://doi.org/10.13039/100014717
GroupedDBID ---
--K
--M
.DC
.~1
0R~
123
1B1
1~.
1~5
4.4
457
4G.
53G
5VS
7-5
71M
8P~
9JM
9JN
AABNK
AAEDT
AAEDW
AAIKJ
AAKOC
AALRI
AAOAW
AAQFI
AATTM
AAXKI
AAXLA
AAXUO
AAYFN
AAYWO
ABBOA
ABCQJ
ABFNM
ABJNI
ABMAC
ACDAQ
ACGFS
ACLOT
ACRLP
ACVFH
ACZNC
ADBBV
ADCNI
ADEZE
AEBSH
AEIPS
AEKER
AENEX
AEUPX
AFJKZ
AFPUW
AFTJW
AFXIZ
AGHFR
AGUBO
AGWIK
AGYEJ
AHHHB
AHZHX
AIALX
AIEXJ
AIGII
AIIUN
AIKHN
AITUG
AKBMS
AKRWK
AKYEP
ALMA_UNASSIGNED_HOLDINGS
AMRAJ
ANKPU
AOUOD
APXCP
AXJTR
BKOJK
BLXMC
CS3
DU5
EBS
EFJIC
EFKBS
EFLBG
EO8
EO9
EP2
EP3
F5P
FDB
FIRID
FNPLU
FYGXN
G-Q
GBLVA
GBOLZ
IHE
J1W
KOM
LG9
M41
MO0
MOBAO
N9A
O-L
O9-
OAUVE
OZT
P-8
P-9
P2P
PC.
Q38
ROL
RPZ
SDF
SDG
SDP
SES
SEW
SPC
SPCBC
SSN
SSV
SSZ
T5K
ZMT
~G-
~HD
29N
9DU
AAQXK
AAYXX
ABWVN
ABXDB
ACNNM
ACRPL
ADJOM
ADMUD
ADNMO
AGQPQ
ASPBG
AVWKF
AZFZN
CITATION
EJD
FEDTE
FGOYB
HLZ
HVGLF
HZ~
R2-
SBC
WUQ
XPP
ID FETCH-LOGICAL-c257t-a0166c7dedbdc3b172222fe79e3615284da4cdd0b8d2e68095face3b44d2b9ba3
ISICitedReferencesCount 1
ISICitedReferencesURI http://www.webofscience.com/api/gateway?GWVersion=2&SrcApp=Summon&SrcAuth=ProQuest&DestLinkType=CitingArticles&DestApp=WOS_CPL&KeyUT=001582098000002&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
ISSN 0925-2312
IngestDate Thu Nov 27 00:53:55 EST 2025
Sat Nov 22 16:51:05 EST 2025
IsPeerReviewed true
IsScholarly true
Keywords Human skeleton action recognition
Spatio-temporal decoupling
Explainability algorithm
Spatio temporal graph convolutional networks
Language English
LinkModel OpenURL
MergedId FETCHMERGED-LOGICAL-c257t-a0166c7dedbdc3b172222fe79e3615284da4cdd0b8d2e68095face3b44d2b9ba3
ORCID 0000-0002-8814-4628
ParticipantIDs crossref_primary_10_1016_j_neucom_2025_131539
elsevier_sciencedirect_doi_10_1016_j_neucom_2025_131539
PublicationCentury 2000
PublicationDate 2025-12-07
PublicationDateYYYYMMDD 2025-12-07
PublicationDate_xml – month: 12
  year: 2025
  text: 2025-12-07
  day: 07
PublicationDecade 2020
PublicationTitle Neurocomputing (Amsterdam)
PublicationYear 2025
Publisher Elsevier B.V
Publisher_xml – name: Elsevier B.V
References Liu, Li, Xu, Han (bib0015) 2025; 133
Wang, Zhang, Zhou, Cui, Yang (bib0005) 2023; 25
Chen, Han, Debattista (bib0060) 2024; 46
Lucic, Hoeve, Tolomei, de Rijke, Silvestri (bib0085) 2022; vol. 151
Zhang, Zhang, Debattista, Han (bib0045) 2025
Schwarzenberg, Hübner, Harbecke, Alt, Hennig (bib0130) 2019
Sun, Ke, Rahmani, Bennamoun, Wang, Liu (bib0040) 2023; 45
Luo, Cheng, Xu, Yu, Zong, Chen, Zhang (bib0105) 2020
Xia, Li, Luo (bib0020) 2022; 24
Li, Xia, Liu (bib0025) 2020; 103
Shao, Han, Marnerides, Debattista (bib0010) 2022; 36
M.S. Schlichtkrull, N.D. Cao, I. Titov, Interpreting graph neural networks for NLP with differentiable edge masking, (2021).
Schnake, Eberle, Lederer, Nakajima, Schütt, Müller, Montavon (bib0125) 2022; 44
Xie, Liu, Shen (bib0165) 2022
Li, Xia, Liu (bib0190) 2020; 103
Fan, Yao, Joe-Wong (bib0160) 2021
Li, Xie, Zhang, Han, Zhen, Chen (bib0065) 2022; 33
Li, Shi, Chen, Zhang, Xie (bib0180) 2024; 46
Zhang, DeFazio, Ramesh (bib0140) 2021
Wang, Luo, Cheng, Chen, Zhang (bib0170) 2025; 19
Xia, Li, Luo (bib0195) 2021; 24
Baldassarre, Azizpour (bib0095) 2019; abs/1905.13686
Pareek, Thakkar (bib0035) 2021; 54
Funke, Khosla, Rathee, Anand (bib0115) 2023
Lee, Lee, Lee, Lee (bib0205) 2023
Shahroudy, Liu, Ng, Wang (bib0210) 2016
Somarathna, Bednarz, Mohammadi (bib0055) 2023; 14
Liu, Zhang, Zhang, Han (bib0070) 2021; 44
Yu, Li, Liang, Chen (bib0175) 2024; 605
Pope, Kolouri, Rostami, Martin, Hoffmann (bib0090) 2019
Altieri, Ceci, Corizzo (bib0200) 2023
Cheng, Zhou, Tay, Wen (bib0150) 2023; 25
Li, Shi, Chen, Zhang, Xie (bib0155) 2024
Liu, Cheng, Zhang, Xu, Han (bib0075) 2024
Ying, Bourgeois, You, Zitnik, Leskovec (bib0100) 2019
Huang, Yamada, Tian, Singh, Chang (bib0145) 2023; 35
Yu, Li, Shi, Xie (bib0185) 2025
Liu, Shahroudy, Perez, Wang, Duan, Kot (bib0215) 2020; 42
Li, Liang, Zheng, Yu (bib0225) 2024; 149
Li, Guo, Liu, Luo, Xie (bib0030) 2022; 31
Yan, Xiong, Lin (bib0080) 2018
Yuan, Yu, Wang, Li, Ji (bib0120) 2021; vol. 139
Vu, Thai (bib0135) 2020
Desai, Ramaswamy (bib0220) 2020
Nakamura (bib0050) 2024; abs/2404.02624
Xia (10.1016/j.neucom.2025.131539_bib0195) 2021; 24
Somarathna (10.1016/j.neucom.2025.131539_bib0055) 2023; 14
Yan (10.1016/j.neucom.2025.131539_bib0080) 2018
Schwarzenberg (10.1016/j.neucom.2025.131539_bib0130) 2019
Wang (10.1016/j.neucom.2025.131539_bib0005) 2023; 25
10.1016/j.neucom.2025.131539_bib0110
Sun (10.1016/j.neucom.2025.131539_bib0040) 2023; 45
Cheng (10.1016/j.neucom.2025.131539_bib0150) 2023; 25
Lee (10.1016/j.neucom.2025.131539_bib0205) 2023
Zhang (10.1016/j.neucom.2025.131539_bib0045) 2025
Vu (10.1016/j.neucom.2025.131539_bib0135) 2020
Zhang (10.1016/j.neucom.2025.131539_bib0140) 2021
Huang (10.1016/j.neucom.2025.131539_bib0145) 2023; 35
Chen (10.1016/j.neucom.2025.131539_bib0060) 2024; 46
Yu (10.1016/j.neucom.2025.131539_bib0175) 2024; 605
Shahroudy (10.1016/j.neucom.2025.131539_bib0210) 2016
Liu (10.1016/j.neucom.2025.131539_bib0215) 2020; 42
Li (10.1016/j.neucom.2025.131539_bib0065) 2022; 33
Schnake (10.1016/j.neucom.2025.131539_bib0125) 2022; 44
Luo (10.1016/j.neucom.2025.131539_bib0105) 2020
Pope (10.1016/j.neucom.2025.131539_bib0090) 2019
Xie (10.1016/j.neucom.2025.131539_bib0165)
Li (10.1016/j.neucom.2025.131539_bib0030) 2022; 31
Li (10.1016/j.neucom.2025.131539_bib0155) 2024
Liu (10.1016/j.neucom.2025.131539_bib0075) 2024
Yu (10.1016/j.neucom.2025.131539_bib0185) 2025
Liu (10.1016/j.neucom.2025.131539_bib0015) 2025; 133
Altieri (10.1016/j.neucom.2025.131539_bib0200) 2023
Funke (10.1016/j.neucom.2025.131539_bib0115) 2023
Liu (10.1016/j.neucom.2025.131539_bib0070) 2021; 44
Fan (10.1016/j.neucom.2025.131539_bib0160) 2021
Desai (10.1016/j.neucom.2025.131539_bib0220) 2020
Xia (10.1016/j.neucom.2025.131539_bib0020) 2022; 24
Li (10.1016/j.neucom.2025.131539_bib0025) 2020; 103
Nakamura (10.1016/j.neucom.2025.131539_bib0050) 2024; abs/2404.02624
Li (10.1016/j.neucom.2025.131539_bib0225) 2024; 149
Ying (10.1016/j.neucom.2025.131539_bib0100) 2019
Li (10.1016/j.neucom.2025.131539_bib0190) 2020; 103
Yuan (10.1016/j.neucom.2025.131539_bib0120) 2021; vol. 139
Wang (10.1016/j.neucom.2025.131539_bib0170) 2025; 19
Shao (10.1016/j.neucom.2025.131539_bib0010) 2022; 36
Pareek (10.1016/j.neucom.2025.131539_bib0035) 2021; 54
Li (10.1016/j.neucom.2025.131539_bib0180) 2024; 46
Lucic (10.1016/j.neucom.2025.131539_bib0085) 2022; vol. 151
Baldassarre (10.1016/j.neucom.2025.131539_bib0095) 2019; abs/1905.13686
References_xml – volume: 36
  start-page: 4184
  year: 2022
  end-page: 4195
  ident: bib0010
  article-title: Region-object relation-aware dense captioning via transformer
  publication-title: IEEE Trans. Neural Netw. Learn. Syst.
– volume: abs/1905.13686
  year: 2019
  ident: bib0095
  article-title: Explainability techniques for graph convolutional networks
  publication-title: CoRr
– volume: 25
  start-page: 8225
  year: 2023
  end-page: 8239
  ident: bib0150
  article-title: Graph neural networks with triple attention for few-shot learning
  publication-title: IEEE Trans. Multim.
– volume: 44
  start-page: 3688
  year: 2021
  end-page: 3704
  ident: bib0070
  article-title: Part-object relational visual saliency
  publication-title: IEEE Trans. Pattern Anal. Mach. Intell.
– start-page: 7444
  year: 2018
  end-page: 7452
  ident: bib0080
  article-title: Spatial temporal graph convolutional networks for skeleton-based action recognition
  publication-title: Thirty-Second AAAI Conference on Artificial Intelligence
– start-page: 1
  year: 2024
  end-page: 14
  ident: bib0155
  article-title: Gt-cam: game theory based class activation map for GCN
  publication-title: IEEE Trans. Pattern Anal. Mach. Intell.
– volume: 14
  start-page: 2626
  year: 2023
  end-page: 2645
  ident: bib0055
  article-title: Virtual reality for emotion elicitation - a review
  publication-title: IEEE Trans. Affect. Comput.
– year: 2025
  ident: bib0185
  article-title: Explaining spatio-temporal graph convolutional networks with spatio-temporal constraints perturbation for action recognition
  publication-title: Inf. Fusion
– reference: M.S. Schlichtkrull, N.D. Cao, I. Titov, Interpreting graph neural networks for NLP with differentiable edge masking, (2021).
– volume: 33
  start-page: 4800
  year: 2022
  end-page: 4814
  ident: bib0065
  article-title: Memory attention networks for skeleton-based action recognition
  publication-title: IEEE Trans. Neural Networks Learn. Syst.
– start-page: 9240
  year: 2019
  end-page: 9251
  ident: bib0100
  article-title: Gnnexplainer: generating explanations for graph neural networks
  publication-title: Advances in Neural Information Processing Systems 32: Annual Conference on Neural Information Processing Systems 2019, NEURIPS 2019
– volume: 42
  start-page: 2684
  year: 2020
  end-page: 2701
  ident: bib0215
  article-title: Ntu RGB+d 120: a large-scale benchmark for 3d human activity understanding
  publication-title: IEEE Trans. Pattern Anal. Mach. Intell.
– year: 2020
  ident: bib0135
  article-title: pgm-explainer: probabilistic graphical model explanations for graph neural networks
  publication-title: Advances in Neural Information Processing Systems 33: Annual Conference on Neural Information Processing Systems 2020, NEURIPS
– volume: 54
  start-page: 2259
  year: 2021
  end-page: 2322
  ident: bib0035
  article-title: A survey on video-based human action recognition: recent updates, datasets, challenges, and applications
  publication-title: Artif. Intell. Rev.
– start-page: 10444
  year: 2023
  end-page: 10453
  ident: bib0205
  article-title: Hierarchically decomposed graph convolutional networks for skeleton-based action recognition
  publication-title: Proceedings of the IEEE/CVF International Conference on Computer Vision
– volume: 133
  start-page: 4483
  year: 2025
  end-page: 4503
  ident: bib0015
  article-title: Part-whole relational fusion towards multi-modal scene understanding
  publication-title: Int. J. Comput. Vis.
– volume: 24
  start-page: 2648
  year: 2022
  end-page: 2661
  ident: bib0020
  article-title: laga-net: local-and-global attention network for skeleton based action recognition
  publication-title: IEEE Trans. Multim.
– volume: 46
  start-page: 5595
  year: 2024
  end-page: 5611
  ident: bib0060
  article-title: Virtual category learning: a semi-supervised learning method for dense prediction with extremely limited labels
  publication-title: IEEE Trans. Pattern Anal. Mach. Intell.
– start-page: 1060
  year: 2021
  end-page: 1065
  ident: bib0160
  article-title: gcn-se: attention as explainability for node classification in dynamic graphs
  publication-title: 2021 IEEE International Conference on Data Mining (ICDM)
– start-page: 174
  year: 2023
  end-page: 188
  ident: bib0200
  article-title: Explainable spatio-temporal graph modeling
  publication-title: International Conference on Discovery Science
– volume: abs/2404.02624
  year: 2024
  ident: bib0050
  article-title: Multi-scale spatial-temporal self-attention graph convolutional networks for skeleton-based action recognition
  publication-title: CoRr
– start-page: 1010
  year: 2016
  end-page: 1019
  ident: bib0210
  article-title: Ntu RGB+d: a large scale dataset for 3d human activity analysis
  publication-title: 2016 IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2016
– volume: vol. 151
  start-page: 4499
  year: 2022
  end-page: 4511
  ident: bib0085
  article-title: Cf-GNNExplainer: counterfactual explanations for graph neural networks
  publication-title: International Conference on Artificial Intelligence and Statistics, AISTATS 2022
– volume: 35
  start-page: 6968
  year: 2023
  end-page: 6972
  ident: bib0145
  article-title: Graphlime: local interpretable model explanations for graph neural networks
  publication-title: IEEE Trans. Knowl. Data Eng.
– volume: 31
  start-page: 168
  year: 2022
  end-page: 180
  ident: bib0030
  article-title: Action status based novel relative feature representations for interaction recognition
  publication-title: Chin. J. Electron.
– volume: 19
  start-page: 1
  year: 2025
  end-page: 21
  ident: bib0170
  article-title: dyexplainer: self-explainable dynamic graph neural network with sparse attentions
  publication-title: ACM Trans. Knowl. Discov. Data
– volume: 103
  year: 2020
  ident: bib0190
  article-title: Learning shape and motion representations for view invariant skeleton-based action recognition
  publication-title: Pattern Recognit.
– year: 2020
  ident: bib0105
  article-title: Parameterized explainer for graph neural network
  publication-title: Advances in Neural Information Processing Systems 33: Annual Conference on Neural Information Processing Systems 2020, NEURIPS 2020
– start-page: 8687
  year: 2023
  end-page: 8698
  ident: bib0115
  article-title: Zorro: valid, sparse, and stable explanations in graph neural networks
  publication-title: IEEE Trans. Knowl. Data Eng.
– year: 2024
  ident: bib0075
  article-title: Capsule networks with residual pose routing
  publication-title: IEEE Trans. Neural Netw. Learn. Syst.
– volume: vol. 139
  start-page: 12241
  year: 2021
  end-page: 12252
  ident: bib0120
  article-title: On explainability of graph neural networks via subgraph explorations
  publication-title: Proceedings of the 38th International Conference on Machine Learning, ICML 2021
– year: 2022
  ident: bib0165
  article-title: Explaining dynamic graph neural networks via relevance back-propagation
– volume: 103
  year: 2020
  ident: bib0025
  article-title: Learning shape and motion representations for view invariant skeleton-based action recognition
  publication-title: Pattern Recognit.
– volume: 25
  start-page: 90
  year: 2023
  end-page: 99
  ident: bib0005
  article-title: Instance-aware deep graph learning for multi-label classification
  publication-title: IEEE Trans. Multim.
– start-page: 58
  year: 2019
  end-page: 62
  ident: bib0130
  article-title: Layerwise relevance visualization in convolutional text graph classifiers
  publication-title: Proceedings of the Thirteenth Workshop on Graph-Based Methods for Natural Language Processing
– volume: 45
  start-page: 3200
  year: 2023
  end-page: 3225
  ident: bib0040
  article-title: Human action recognition from various data modalities: a review
  publication-title: IEEE Trans. Pattern Anal. Mach. Intell.
– volume: 44
  start-page: 7581
  year: 2022
  end-page: 7596
  ident: bib0125
  article-title: Higher-order explanations of graph neural networks via relevant walks
  publication-title: IEEE Trans. Pattern Anal. Mach. Intell.
– volume: 24
  start-page: 2648
  year: 2021
  end-page: 2661
  ident: bib0195
  article-title: laga-net: local-and-global attention network for skeleton based action recognition
  publication-title: IEEE Trans. Multimed.
– volume: 46
  start-page: 8806
  year: 2024
  end-page: 8819
  ident: bib0180
  article-title: Gt-cam: game theory based class activation map for GCN
  publication-title: IEEE Trans. Pattern Anal. Mach. Intell.
– volume: 605
  year: 2024
  ident: bib0175
  article-title: Geoexplainer: interpreting graph convolutional networks with geometric masking
  publication-title: Neurocomputing
– start-page: 1042
  year: 2021
  end-page: 1049
  ident: bib0140
  article-title: relex: a model-agnostic relational model explainer
  publication-title: AIES ’21: AAAI/ACM Conference on AI, Ethics, and Society, Virtual Event
– volume: 149
  year: 2024
  ident: bib0225
  article-title: cr-cam: generating explanations for deep neural networks by contrasting and ranking features
  publication-title: Pattern Recognit.
– start-page: 10772
  year: 2019
  end-page: 10781
  ident: bib0090
  article-title: Explainability methods for graph convolutional neural networks
  publication-title: IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2019
– year: 2025
  ident: bib0045
  article-title: Cross-modality distillation for multi-modal tracking
  publication-title: IEEE Trans. Pattern Anal. Mach. Intell.
– start-page: 972
  year: 2020
  end-page: 980
  ident: bib0220
  article-title: Ablation-cam: visual explanations for deep convolutional network via gradient-free localization
  publication-title: IEEE Winter Conference on Applications of Computer Vision, WACV 2020
– volume: 33
  start-page: 4800
  year: 2022
  ident: 10.1016/j.neucom.2025.131539_bib0065
  article-title: Memory attention networks for skeleton-based action recognition
  publication-title: IEEE Trans. Neural Networks Learn. Syst.
  doi: 10.1109/TNNLS.2021.3061115
– ident: 10.1016/j.neucom.2025.131539_bib0165
– year: 2025
  ident: 10.1016/j.neucom.2025.131539_bib0045
  article-title: Cross-modality distillation for multi-modal tracking
  publication-title: IEEE Trans. Pattern Anal. Mach. Intell.
– start-page: 972
  year: 2020
  ident: 10.1016/j.neucom.2025.131539_bib0220
  article-title: Ablation-cam: visual explanations for deep convolutional network via gradient-free localization
– start-page: 1060
  year: 2021
  ident: 10.1016/j.neucom.2025.131539_bib0160
  article-title: gcn-se: attention as explainability for node classification in dynamic graphs
– year: 2025
  ident: 10.1016/j.neucom.2025.131539_bib0185
  article-title: Explaining spatio-temporal graph convolutional networks with spatio-temporal constraints perturbation for action recognition
  publication-title: Inf. Fusion
  doi: 10.1016/j.inffus.2025.103387
– volume: vol. 151
  start-page: 4499
  year: 2022
  ident: 10.1016/j.neucom.2025.131539_bib0085
  article-title: Cf-GNNExplainer: counterfactual explanations for graph neural networks
– volume: 19
  start-page: 1
  year: 2025
  ident: 10.1016/j.neucom.2025.131539_bib0170
  article-title: dyexplainer: self-explainable dynamic graph neural network with sparse attentions
  publication-title: ACM Trans. Knowl. Discov. Data
– volume: 25
  start-page: 90
  year: 2023
  ident: 10.1016/j.neucom.2025.131539_bib0005
  article-title: Instance-aware deep graph learning for multi-label classification
  publication-title: IEEE Trans. Multim.
  doi: 10.1109/TMM.2021.3121559
– volume: 31
  start-page: 168
  year: 2022
  ident: 10.1016/j.neucom.2025.131539_bib0030
  article-title: Action status based novel relative feature representations for interaction recognition
  publication-title: Chin. J. Electron.
– volume: abs/2404.02624
  year: 2024
  ident: 10.1016/j.neucom.2025.131539_bib0050
  article-title: Multi-scale spatial-temporal self-attention graph convolutional networks for skeleton-based action recognition
  publication-title: CoRr
– volume: 24
  start-page: 2648
  year: 2022
  ident: 10.1016/j.neucom.2025.131539_bib0020
  article-title: laga-net: local-and-global attention network for skeleton based action recognition
  publication-title: IEEE Trans. Multim.
  doi: 10.1109/TMM.2021.3086758
– volume: 133
  start-page: 4483
  year: 2025
  ident: 10.1016/j.neucom.2025.131539_bib0015
  article-title: Part-whole relational fusion towards multi-modal scene understanding
  publication-title: Int. J. Comput. Vis.
  doi: 10.1007/s11263-025-02393-8
– volume: 44
  start-page: 7581
  year: 2022
  ident: 10.1016/j.neucom.2025.131539_bib0125
  article-title: Higher-order explanations of graph neural networks via relevant walks
  publication-title: IEEE Trans. Pattern Anal. Mach. Intell.
  doi: 10.1109/TPAMI.2021.3115452
– start-page: 1010
  year: 2016
  ident: 10.1016/j.neucom.2025.131539_bib0210
  article-title: Ntu RGB+d: a large scale dataset for 3d human activity analysis
– start-page: 10772
  year: 2019
  ident: 10.1016/j.neucom.2025.131539_bib0090
  article-title: Explainability methods for graph convolutional neural networks
– volume: 42
  start-page: 2684
  year: 2020
  ident: 10.1016/j.neucom.2025.131539_bib0215
  article-title: Ntu RGB+d 120: a large-scale benchmark for 3d human activity understanding
  publication-title: IEEE Trans. Pattern Anal. Mach. Intell.
  doi: 10.1109/TPAMI.2019.2916873
– year: 2020
  ident: 10.1016/j.neucom.2025.131539_bib0135
  article-title: pgm-explainer: probabilistic graphical model explanations for graph neural networks
– volume: 35
  start-page: 6968
  year: 2023
  ident: 10.1016/j.neucom.2025.131539_bib0145
  article-title: Graphlime: local interpretable model explanations for graph neural networks
  publication-title: IEEE Trans. Knowl. Data Eng.
  doi: 10.1109/TKDE.2022.3187455
– volume: 25
  start-page: 8225
  year: 2023
  ident: 10.1016/j.neucom.2025.131539_bib0150
  article-title: Graph neural networks with triple attention for few-shot learning
  publication-title: IEEE Trans. Multim.
  doi: 10.1109/TMM.2022.3233442
– ident: 10.1016/j.neucom.2025.131539_bib0110
– volume: 103
  year: 2020
  ident: 10.1016/j.neucom.2025.131539_bib0190
  article-title: Learning shape and motion representations for view invariant skeleton-based action recognition
  publication-title: Pattern Recognit.
  doi: 10.1016/j.patcog.2020.107293
– volume: 24
  start-page: 2648
  year: 2021
  ident: 10.1016/j.neucom.2025.131539_bib0195
  article-title: laga-net: local-and-global attention network for skeleton based action recognition
  publication-title: IEEE Trans. Multimed.
  doi: 10.1109/TMM.2021.3086758
– start-page: 174
  year: 2023
  ident: 10.1016/j.neucom.2025.131539_bib0200
  article-title: Explainable spatio-temporal graph modeling
– volume: abs/1905.13686
  year: 2019
  ident: 10.1016/j.neucom.2025.131539_bib0095
  article-title: Explainability techniques for graph convolutional networks
  publication-title: CoRr
– volume: 14
  start-page: 2626
  year: 2023
  ident: 10.1016/j.neucom.2025.131539_bib0055
  article-title: Virtual reality for emotion elicitation - a review
  publication-title: IEEE Trans. Affect. Comput.
  doi: 10.1109/TAFFC.2022.3181053
– start-page: 9240
  year: 2019
  ident: 10.1016/j.neucom.2025.131539_bib0100
  article-title: Gnnexplainer: generating explanations for graph neural networks
– volume: 46
  start-page: 8806
  year: 2024
  ident: 10.1016/j.neucom.2025.131539_bib0180
  article-title: Gt-cam: game theory based class activation map for GCN
  publication-title: IEEE Trans. Pattern Anal. Mach. Intell.
  doi: 10.1109/TPAMI.2024.3413026
– volume: 46
  start-page: 5595
  year: 2024
  ident: 10.1016/j.neucom.2025.131539_bib0060
  article-title: Virtual category learning: a semi-supervised learning method for dense prediction with extremely limited labels
  publication-title: IEEE Trans. Pattern Anal. Mach. Intell.
  doi: 10.1109/TPAMI.2024.3367416
– year: 2024
  ident: 10.1016/j.neucom.2025.131539_bib0075
  article-title: Capsule networks with residual pose routing
  publication-title: IEEE Trans. Neural Netw. Learn. Syst.
– volume: 149
  year: 2024
  ident: 10.1016/j.neucom.2025.131539_bib0225
  article-title: cr-cam: generating explanations for deep neural networks by contrasting and ranking features
  publication-title: Pattern Recognit.
  doi: 10.1016/j.patcog.2024.110251
– start-page: 1
  year: 2024
  ident: 10.1016/j.neucom.2025.131539_bib0155
  article-title: Gt-cam: game theory based class activation map for GCN
  publication-title: IEEE Trans. Pattern Anal. Mach. Intell.
– volume: 45
  start-page: 3200
  year: 2023
  ident: 10.1016/j.neucom.2025.131539_bib0040
  article-title: Human action recognition from various data modalities: a review
  publication-title: IEEE Trans. Pattern Anal. Mach. Intell.
– year: 2020
  ident: 10.1016/j.neucom.2025.131539_bib0105
  article-title: Parameterized explainer for graph neural network
– start-page: 58
  year: 2019
  ident: 10.1016/j.neucom.2025.131539_bib0130
  article-title: Layerwise relevance visualization in convolutional text graph classifiers
– start-page: 10444
  year: 2023
  ident: 10.1016/j.neucom.2025.131539_bib0205
  article-title: Hierarchically decomposed graph convolutional networks for skeleton-based action recognition
– volume: 44
  start-page: 3688
  year: 2021
  ident: 10.1016/j.neucom.2025.131539_bib0070
  article-title: Part-object relational visual saliency
  publication-title: IEEE Trans. Pattern Anal. Mach. Intell.
– start-page: 1042
  year: 2021
  ident: 10.1016/j.neucom.2025.131539_bib0140
  article-title: relex: a model-agnostic relational model explainer
– volume: 36
  start-page: 4184
  year: 2022
  ident: 10.1016/j.neucom.2025.131539_bib0010
  article-title: Region-object relation-aware dense captioning via transformer
  publication-title: IEEE Trans. Neural Netw. Learn. Syst.
  doi: 10.1109/TNNLS.2022.3152990
– volume: 54
  start-page: 2259
  year: 2021
  ident: 10.1016/j.neucom.2025.131539_bib0035
  article-title: A survey on video-based human action recognition: recent updates, datasets, challenges, and applications
  publication-title: Artif. Intell. Rev.
  doi: 10.1007/s10462-020-09904-8
– start-page: 8687
  year: 2023
  ident: 10.1016/j.neucom.2025.131539_bib0115
  article-title: Zorro: valid, sparse, and stable explanations in graph neural networks
  publication-title: IEEE Trans. Knowl. Data Eng.
  doi: 10.1109/TKDE.2022.3201170
– volume: 103
  year: 2020
  ident: 10.1016/j.neucom.2025.131539_bib0025
  article-title: Learning shape and motion representations for view invariant skeleton-based action recognition
  publication-title: Pattern Recognit.
  doi: 10.1016/j.patcog.2020.107293
– volume: vol. 139
  start-page: 12241
  year: 2021
  ident: 10.1016/j.neucom.2025.131539_bib0120
  article-title: On explainability of graph neural networks via subgraph explorations
– volume: 605
  year: 2024
  ident: 10.1016/j.neucom.2025.131539_bib0175
  article-title: Geoexplainer: interpreting graph convolutional networks with geometric masking
  publication-title: Neurocomputing
  doi: 10.1016/j.neucom.2024.128393
– start-page: 7444
  year: 2018
  ident: 10.1016/j.neucom.2025.131539_bib0080
  article-title: Spatial temporal graph convolutional networks for skeleton-based action recognition
SSID ssj0017129
Score 2.4647171
Snippet Spatio-temporal graph convolutional networks utilize an alternating combination of one-dimensional ordinary convolution and graph convolution to extract...
SourceID crossref
elsevier
SourceType Index Database
Publisher
StartPage 131539
SubjectTerms Explainability algorithm
Human skeleton action recognition
Spatio temporal graph convolutional networks
Spatio-temporal decoupling
Title STD-Explain: Generalizing explanations for spatio-temporal graph convolutional networks based on spatio-temporal decoupled perturbation
URI https://dx.doi.org/10.1016/j.neucom.2025.131539
Volume 657
WOSCitedRecordID wos001582098000002&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: PRVESC
  databaseName: Elsevier SD Freedom Collection Journals 2021
  issn: 0925-2312
  databaseCode: AIEXJ
  dateStart: 19950101
  customDbUrl:
  isFulltext: true
  dateEnd: 99991231
  titleUrlDefault: https://www.sciencedirect.com
  omitProxy: false
  ssIdentifier: ssj0017129
  providerName: Elsevier
link http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwtV1Lb9NAEF6FlgMXKC9RoGgP3CpX8fqxNreoLQKEKqQEKXCx9hU1UbqNkrgK_AHEv-7sy3YbhAAJHyxr4l1bO19mZsffziL0OhMiUyxlUcbkJEozxiKeEwr_q8wUyKJ5n3K72QQ9OyvG4_JTr_czrIW5mlOti82mXPxXVYMMlG2Wzv6FuptOQQDXoHQ4g9rh_EeKH45OIsOsM1N-mO77utLT7yYnoIxce_abIRiuLKE68gWq5oe2frWlovt3BJl2TPHVofF40nxduN1Kwhy2Xszhx4Vagg_jrbpnoTpUDZ7S7iDhcxODC1OiQRo8NrmIj5Za8AW853kL2qHddhhA5X2sTdw6QtF0U7f3Hft1Jl_Pp9-C2KczSGapIbTNsW2ts3HJSrgRItEbdjt3la23fIBLR8yOtKoNIcg85ChOwLKXrc9rmIhD07XpGUJBApPh4g7aJTQrwcbvDt6fjj80n6RoTFzhRv8qYR2mJQtuP-vXcU4ndhntoft-0oEHDiwPUU_pR-hB2NADe_v-GP3oYOcN7iIHd5GDATn4FgawRQ6-gRwckIMtcvCl3mrVIAd3kfMEfX57Ojp-F_mdOiIBJn8dMRiFXFCpJJci4RAUwzFRtFQJRMwQAUmWCin7vJBE5QWE9RMmVMLTVBJecpY8RTv6UqtnCJOSsImR5SxOwcyVeRyXAsZfJLRPaLyPojCu1cIVZKkCU3FWOT1URg-V08M-omHwKx9UumCxArz8tuXzf275At1rof0S7ayXtTpAd8XVerpavvLAugZm3aZd
linkProvider Elsevier
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=STD-Explain%3A+Generalizing+explanations+for+spatio-temporal+graph+convolutional+networks+based+on+spatio-temporal+decoupled+perturbation&rft.jtitle=Neurocomputing+%28Amsterdam%29&rft.au=Li%2C+Yanshan&rft.au=Shi%2C+Ting&rft.au=He%2C+Suixuan&rft.au=Chen%2C+Zhiyuan&rft.date=2025-12-07&rft.pub=Elsevier+B.V&rft.issn=0925-2312&rft.volume=657&rft_id=info:doi/10.1016%2Fj.neucom.2025.131539&rft.externalDocID=S0925231225022118
thumbnail_l http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/lc.gif&issn=0925-2312&client=summon
thumbnail_m http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/mc.gif&issn=0925-2312&client=summon
thumbnail_s http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/sc.gif&issn=0925-2312&client=summon