An interpretable unsupervised capsule network via comprehensive contrastive learning and two-stage training
Limited attention has been given to unsupervised capsule networks (CapsNets) with contrastive learning due to the challenge of harmoniously learning interpretable primary and high-level capsules. To address this issue, we focus on three aspects: loss function, routing algorithm, and training strateg...
Uložené v:
| Vydané v: | Pattern recognition Ročník 158; s. 111059 |
|---|---|
| Hlavní autori: | , , |
| Médium: | Journal Article |
| Jazyk: | English |
| Vydavateľské údaje: |
Elsevier Ltd
01.02.2025
|
| Predmet: | |
| ISSN: | 0031-3203 |
| On-line prístup: | Získať plný text |
| Tagy: |
Pridať tag
Žiadne tagy, Buďte prvý, kto otaguje tento záznam!
|
| Abstract | Limited attention has been given to unsupervised capsule networks (CapsNets) with contrastive learning due to the challenge of harmoniously learning interpretable primary and high-level capsules. To address this issue, we focus on three aspects: loss function, routing algorithm, and training strategy. First, we propose a comprehensive contrastive loss to ensure consistency in learning both high-level and primary capsules across different objects. Next, we introduce an agreement-based routing mechanism for the activation of high-level capsules. Finally, we present a two-stage training strategy to resolve conflicts between multiple losses. Ablation experiments show that these methods all improve model performance. Results from linear evaluation and semi-supervised learning demonstrate that our model outperforms other CapsNets and convolutional neural networks in learning high-level capsules. Additionally, visualizing capsules provides insights into the primary capsules, which remain consistent across images and align with human vision.
•Propose unsupervised CapsNet based on contrastive learning.•Develop a comprehensive contrastive loss for interpretable capsules.•Implement a two-stage training strategy to resolve loss function conflicts.•Design an agreement routing algorithm considering prediction similarity and distribution. |
|---|---|
| AbstractList | Limited attention has been given to unsupervised capsule networks (CapsNets) with contrastive learning due to the challenge of harmoniously learning interpretable primary and high-level capsules. To address this issue, we focus on three aspects: loss function, routing algorithm, and training strategy. First, we propose a comprehensive contrastive loss to ensure consistency in learning both high-level and primary capsules across different objects. Next, we introduce an agreement-based routing mechanism for the activation of high-level capsules. Finally, we present a two-stage training strategy to resolve conflicts between multiple losses. Ablation experiments show that these methods all improve model performance. Results from linear evaluation and semi-supervised learning demonstrate that our model outperforms other CapsNets and convolutional neural networks in learning high-level capsules. Additionally, visualizing capsules provides insights into the primary capsules, which remain consistent across images and align with human vision.
•Propose unsupervised CapsNet based on contrastive learning.•Develop a comprehensive contrastive loss for interpretable capsules.•Implement a two-stage training strategy to resolve loss function conflicts.•Design an agreement routing algorithm considering prediction similarity and distribution. |
| ArticleNumber | 111059 |
| Author | Song, Yan Zeng, Ru Zhong, Yanjiu |
| Author_xml | – sequence: 1 givenname: Ru orcidid: 0000-0002-4270-9662 surname: Zeng fullname: Zeng, Ru email: zengru_neo@163.com – sequence: 2 givenname: Yan orcidid: 0000-0002-9035-9142 surname: Song fullname: Song, Yan email: sonya@usst.edu.cn – sequence: 3 givenname: Yanjiu surname: Zhong fullname: Zhong, Yanjiu email: zhongyanjiu721@163.com |
| BookMark | eNqFkE1OwzAQhb0oEm3hBix8gQQ7dpqaBVJV8SchsYG15djj4jZ1ItsN4vYkhBULWM3Mm_meNG-BZr71gNAVJTkldHW9zzuVdLvLC1LwnFJKSjFDc0IYzVhB2DlaxLgnhFaUF3N02HjsfILQBUiqbgCffDx1EHoXwWCtungaRA_pow0H3DuFdXscjt_BR9fDMPkUVExj34AK3vkdVt7gAchiUjvAw96N8gU6s6qJcPlTl-jt_u51-5g9vzw8bTfPmWZklTJmDSsElLou10QIQSw3qrKcC15aoapKFMIYIQpbEbUymjOr1oPOrKa6rmq2RHzy1aGNMYCVXXBHFT4lJXIMSe7lFJIcQ5JTSAN28wvTLqnkvh90zX_w7QTD8FjvIMioHXgNxgXQSZrW_W3wBfAtjQY |
| CitedBy_id | crossref_primary_10_1371_journal_pone_0329202 |
| Cites_doi | 10.1016/j.eswa.2023.122284 10.1109/CVPR46437.2021.01408 10.1109/CVPR42600.2020.00975 10.1109/ICCV.2019.00580 10.1007/978-3-642-21735-7_6 10.1016/j.patcog.2021.108486 10.1109/CVPR.2017.634 10.1162/neco_a_01557 10.1038/s41598-021-93977-0 10.1016/j.patcog.2023.110142 10.1109/TII.2021.3128412 10.1109/CVPR.2004.1315150 10.1609/aaai.v38i20.30228 10.1109/5.726791 10.1016/j.patcog.2022.109270 10.1109/CVPR.2016.90 10.1016/j.neucom.2023.126916 |
| ContentType | Journal Article |
| Copyright | 2024 Elsevier Ltd |
| Copyright_xml | – notice: 2024 Elsevier Ltd |
| DBID | AAYXX CITATION |
| DOI | 10.1016/j.patcog.2024.111059 |
| DatabaseName | CrossRef |
| DatabaseTitle | CrossRef |
| DatabaseTitleList | |
| DeliveryMethod | fulltext_linktorsrc |
| Discipline | Computer Science |
| ExternalDocumentID | 10_1016_j_patcog_2024_111059 S0031320324008100 |
| GroupedDBID | --K --M -D8 -DT -~X .DC .~1 0R~ 123 1B1 1RT 1~. 1~5 29O 4.4 457 4G. 53G 5VS 7-5 71M 8P~ 9JN AABNK AACTN AAEDT AAEDW AAIKJ AAKOC AALRI AAOAW AAQFI AAQXK AAXKI AAXUO AAYFN ABBOA ABDPE ABEFU ABFNM ABFRF ABHFT ABJNI ABMAC ABTAH ABXDB ACBEA ACDAQ ACGFO ACGFS ACNNM ACRLP ACZNC ADBBV ADEZE ADJOM ADMUD ADMXK ADTZH AEBSH AECPX AEFWE AEKER AENEX AFJKZ AFKWA AFTJW AGHFR AGUBO AGYEJ AHHHB AHJVU AHZHX AIALX AIEXJ AIKHN AITUG AJOXV AKRWK ALMA_UNASSIGNED_HOLDINGS AMFUW AMRAJ AOUOD ASPBG AVWKF AXJTR AZFZN BJAXD BKOJK BLXMC CS3 DU5 EBS EFJIC EJD EO8 EO9 EP2 EP3 F0J F5P FD6 FDB FEDTE FGOYB FIRID FNPLU FYGXN G-Q GBLVA GBOLZ HLZ HVGLF HZ~ H~9 IHE J1W JJJVA KOM KZ1 LG9 LMP LY1 M41 MO0 N9A O-L O9- OAUVE OZT P-8 P-9 P2P PC. Q38 R2- RIG RNS ROL RPZ SBC SDF SDG SDP SDS SES SEW SPC SPCBC SST SSV SSZ T5K TN5 UNMZH VOH WUQ XJE XPP ZMT ZY4 ~G- 9DU AATTM AAYWO AAYXX ABWVN ACLOT ACRPL ACVFH ADCNI ADNMO AEIPS AEUPX AFPUW AGQPQ AIGII AIIUN AKBMS AKYEP ANKPU APXCP CITATION EFKBS EFLBG ~HD |
| ID | FETCH-LOGICAL-c306t-3fd329e5cb5809990f4da7f44945f9a77929dd992f70a6dc43fa89a73fc1cb7b3 |
| ISICitedReferencesCount | 1 |
| ISICitedReferencesURI | http://www.webofscience.com/api/gateway?GWVersion=2&SrcApp=Summon&SrcAuth=ProQuest&DestLinkType=CitingArticles&DestApp=WOS_CPL&KeyUT=001331367700001&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D |
| ISSN | 0031-3203 |
| IngestDate | Tue Nov 18 20:41:47 EST 2025 Sat Nov 29 03:52:41 EST 2025 Sat Nov 16 15:59:03 EST 2024 |
| IsPeerReviewed | true |
| IsScholarly | true |
| Keywords | Unsupervised capsule network Routing algorithm Comprehensive contrastive loss Interpretability Primary capsules |
| Language | English |
| LinkModel | OpenURL |
| MergedId | FETCHMERGED-LOGICAL-c306t-3fd329e5cb5809990f4da7f44945f9a77929dd992f70a6dc43fa89a73fc1cb7b3 |
| ORCID | 0000-0002-9035-9142 0000-0002-4270-9662 |
| ParticipantIDs | crossref_primary_10_1016_j_patcog_2024_111059 crossref_citationtrail_10_1016_j_patcog_2024_111059 elsevier_sciencedirect_doi_10_1016_j_patcog_2024_111059 |
| PublicationCentury | 2000 |
| PublicationDate | February 2025 2025-02-00 |
| PublicationDateYYYYMMDD | 2025-02-01 |
| PublicationDate_xml | – month: 02 year: 2025 text: February 2025 |
| PublicationDecade | 2020 |
| PublicationTitle | Pattern recognition |
| PublicationYear | 2025 |
| Publisher | Elsevier Ltd |
| Publisher_xml | – name: Elsevier Ltd |
| References | H. Liu, C. Li, X. Zhang, et al., Depression Detection via Capsule Networks with Contrastive Learning, in: Proceedings of the AAAI Conference on Artificial Intelligence, Vol. 38, (20) 2024, pp. 22231–22239. J. Gu, V. Tresp, H. Hu, Capsule network is not more robust than convolutional network, in: IEEE Conference on Computer Vision and Pattern Recognition, 2021, pp. 14309–14317. Liu, Dong, Zhang, Xu (b11) 2024; 563 Wah, Branson, Welinder (b39) 2011 Hadian, Zhang, Etesami (b15) 2024 Zeng, Qin, Song (b29) 2024; 238 K. He, X. Zhang, S. Ren, et al., Deep residual learning for image recognition, in: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2016, pp. 770–778. Y. Netzer, T. Wang, A. Coates, et al., Reading digits in natural images with unsupervised feature learning, in: Conference and Workshop on Neural Information Processing Systems No. 5, 2011, p. 7. Cheng, Wang, Wang, Zhang, Han, Zhang (b12) 2023; 137 Shi, Niu, Zhou (b21) 2022; 124 Poličar, Stražar, Zupan (b40) 2019 Hahn, Pyeon, Kim (b37) 2019; 32 K. He, H. Fan, Y. Wu, et al., Momentum contrast for unsupervised visual representation learning, in: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2020, p. 9729. Wang, Wang, Chen (b8) 2024; 141 Liu, Zhang, Zhang, Han (b13) 2021; 44 Y. LeCun, F. Huang, L. Bottou, Learning methods for generic object recognition with invariance to pose and lighting, in: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, Vol. 2, 2004, p. 104. W. Wang, Z. Zhang, S. Qi, et al., Learning compositional neural information fusion for human parsing, in: Proceedings of the IEEE/CVF International Conference on Computer Vision, 2019, pp. 5703–5713. G. Hinton, S. Sabour, N. Frosst, Matrix capsules with EM routing, in: International Conference on Learning Representations, 2018. Kosiorek, Sabour, Teh, Hinton (b4) 2019; 32 Zhang, Zhang, Zhang, Wang, Wang, Zheng (b20) 2024; 12 Chen, Shen, Zhao, Yang, Zhang (b14) 2024; 148 Mazzia, Salvetti, Chiaberge (b28) 2021; 11 Zeng, Song (b22) 2022; 18 Howard, Zhu, Chen (b25) 2017 Hu, Lu, Zhou (b16) 2024; 3 Sun, Tagliasacchi, Sabour, Hinton (b1) 2021; 34 M. Tan, Q. Le, Efficientnet: Rethinking model scaling for convolutional neural networks, in: International Conference on Machine Learning, 2019, pp. 6105–6114. Krizhevsky, Hinton (b31) 2019 Xiao, Rasul, Vollgraf (b33) 2017 Kingma, Ba (b36) 2014 G. Hinton, A. Krizhevsky, S. Wang, Transforming auto-encoders, in: International Conference on Neural Networks, 2011, pp. 44–51. Liu, Cheng, Zhang, Xu, Han (b10) 2024 T. Chen, S. Kornblith, M. Norouzi, G.E. Hinton, A simple framework for contrastive learning of visual representations, in: International Conference on Machine Learning, 2020, pp. 1597–1607. S. Sabour, A. Tagliasacchi, S. Yazdani, G. Hinton, D.J. Fleet, Unsupervised part representation by flow capsules, in: International Conference on Machine Learning, 2021, pp. 9213–9223. Hinton (b2) 2023; 35 Panwar, Patras (b7) 2022 Culp, Sabour, Hinton (b19) 2022 S. Sabour, N. Frosst, G.E. Hinton, Dynamic routing between capsules, in: Conference and Workshop on Neural Information Processing Systems, 2017, pp. 3859–3869. S. Xie, R. Girshick, K. He, et al., Aggregated residual transformations for deep neural networks, in: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2017, pp. 1492–1500. LeCun, Bottou, Bengio, Haffner (b32) 1998; 86 Culp (10.1016/j.patcog.2024.111059_b19) 2022 Hinton (10.1016/j.patcog.2024.111059_b2) 2023; 35 Shi (10.1016/j.patcog.2024.111059_b21) 2022; 124 Chen (10.1016/j.patcog.2024.111059_b14) 2024; 148 10.1016/j.patcog.2024.111059_b18 LeCun (10.1016/j.patcog.2024.111059_b32) 1998; 86 10.1016/j.patcog.2024.111059_b17 Krizhevsky (10.1016/j.patcog.2024.111059_b31) 2019 10.1016/j.patcog.2024.111059_b38 Sun (10.1016/j.patcog.2024.111059_b1) 2021; 34 10.1016/j.patcog.2024.111059_b35 10.1016/j.patcog.2024.111059_b23 Hahn (10.1016/j.patcog.2024.111059_b37) 2019; 32 Panwar (10.1016/j.patcog.2024.111059_b7) 2022 Hadian (10.1016/j.patcog.2024.111059_b15) 2024 Howard (10.1016/j.patcog.2024.111059_b25) 2017 Liu (10.1016/j.patcog.2024.111059_b13) 2021; 44 Hu (10.1016/j.patcog.2024.111059_b16) 2024; 3 Zhang (10.1016/j.patcog.2024.111059_b20) 2024; 12 Zeng (10.1016/j.patcog.2024.111059_b29) 2024; 238 10.1016/j.patcog.2024.111059_b5 Wah (10.1016/j.patcog.2024.111059_b39) 2011 10.1016/j.patcog.2024.111059_b3 10.1016/j.patcog.2024.111059_b9 10.1016/j.patcog.2024.111059_b6 Kingma (10.1016/j.patcog.2024.111059_b36) 2014 10.1016/j.patcog.2024.111059_b27 10.1016/j.patcog.2024.111059_b26 Wang (10.1016/j.patcog.2024.111059_b8) 2024; 141 Liu (10.1016/j.patcog.2024.111059_b10) 2024 Liu (10.1016/j.patcog.2024.111059_b11) 2024; 563 10.1016/j.patcog.2024.111059_b24 10.1016/j.patcog.2024.111059_b34 Xiao (10.1016/j.patcog.2024.111059_b33) 2017 10.1016/j.patcog.2024.111059_b30 Poličar (10.1016/j.patcog.2024.111059_b40) 2019 Kosiorek (10.1016/j.patcog.2024.111059_b4) 2019; 32 Cheng (10.1016/j.patcog.2024.111059_b12) 2023; 137 Mazzia (10.1016/j.patcog.2024.111059_b28) 2021; 11 Zeng (10.1016/j.patcog.2024.111059_b22) 2022; 18 |
| References_xml | – volume: 238 year: 2024 ident: b29 article-title: A non-iterative capsule network with interdependent agreement routing publication-title: Expert Syst. Appl. – volume: 137 year: 2023 ident: b12 article-title: Hybrid routing transformer for zero-shot learning publication-title: Pattern Recognit. – volume: 34 start-page: 24993 year: 2021 end-page: 25005 ident: b1 article-title: Canonical capsules: Self-supervised capsules in canonical pose publication-title: Adv. Neural Inf. Process. Syst. – volume: 44 start-page: 3688 year: 2021 end-page: 3704 ident: b13 article-title: Part-object relational visual saliency publication-title: IEEE Trans. Pattern Anal. Mach. Intell. – year: 2011 ident: b39 article-title: The caltech-ucsd birds-200–2011 dataset – year: 2019 ident: b40 article-title: openTSNE: a modular Python library for t-SNE dimensionality reduction and embedding publication-title: BioRxiv – volume: 32 year: 2019 ident: b4 article-title: Stacked capsule autoencoders publication-title: Adv. Neural Inf. Process. Syst. – reference: K. He, H. Fan, Y. Wu, et al., Momentum contrast for unsupervised visual representation learning, in: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2020, p. 9729. – year: 2019 ident: b31 article-title: Learning Multiple Layers of Features from Tiny Images – volume: 35 start-page: 413 year: 2023 end-page: 452 ident: b2 article-title: How to represent part-whole hierarchies in a neural network publication-title: Neural Comput. – volume: 11 start-page: 14634 year: 2021 ident: b28 article-title: Efficient-capsnet: Capsule network with self-attention routing publication-title: Sci. Rep. – year: 2017 ident: b25 article-title: Mobilenets: Efficient convolutional neural networks for mobile vision applications – volume: 141 year: 2024 ident: b8 article-title: RPI-CapsuleGAN: Predicting RNA-protein interactions through an interpretable generative adversarial capsule network publication-title: Pattern Recognit. – reference: W. Wang, Z. Zhang, S. Qi, et al., Learning compositional neural information fusion for human parsing, in: Proceedings of the IEEE/CVF International Conference on Computer Vision, 2019, pp. 5703–5713. – reference: K. He, X. Zhang, S. Ren, et al., Deep residual learning for image recognition, in: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2016, pp. 770–778. – reference: M. Tan, Q. Le, Efficientnet: Rethinking model scaling for convolutional neural networks, in: International Conference on Machine Learning, 2019, pp. 6105–6114. – volume: 32 start-page: 7658 year: 2019 end-page: 7667 ident: b37 article-title: Self-routing capsule networks publication-title: Adv. Neural Inf. Process. Syst. – reference: G. Hinton, S. Sabour, N. Frosst, Matrix capsules with EM routing, in: International Conference on Learning Representations, 2018. – reference: Y. Netzer, T. Wang, A. Coates, et al., Reading digits in natural images with unsupervised feature learning, in: Conference and Workshop on Neural Information Processing Systems No. 5, 2011, p. 7. – year: 2022 ident: b19 article-title: Testing GLOM’s ability to infer wholes from ambiguous parts – reference: S. Sabour, N. Frosst, G.E. Hinton, Dynamic routing between capsules, in: Conference and Workshop on Neural Information Processing Systems, 2017, pp. 3859–3869. – reference: S. Sabour, A. Tagliasacchi, S. Yazdani, G. Hinton, D.J. Fleet, Unsupervised part representation by flow capsules, in: International Conference on Machine Learning, 2021, pp. 9213–9223. – start-page: 1 year: 2024 end-page: 15 ident: b15 article-title: A robust interpolated model predictive control based on recurrent neural networks for a nonholonomic differential-drive mobile robot with quasi-LPV representation: computational complexity and conservatism publication-title: Int. J. Syst. Sci. – volume: 86 start-page: 2278 year: 1998 end-page: 2324 ident: b32 article-title: Gradient-based learning applied to document recognition publication-title: Proc. IEEE – volume: 124 year: 2022 ident: b21 article-title: Sparse CapsNet with explicit regularizer publication-title: Pattern Recognit. – volume: 18 start-page: 4383 year: 2022 end-page: 4392 ident: b22 article-title: A fast routing capsule network with improved dense blocks publication-title: IEEE Trans. Ind. Inform. – reference: G. Hinton, A. Krizhevsky, S. Wang, Transforming auto-encoders, in: International Conference on Neural Networks, 2011, pp. 44–51. – year: 2017 ident: b33 article-title: Fashion-mnist: a novel image dataset for benchmarking machine learning algorithms – volume: 3 year: 2024 ident: b16 article-title: Learning regression distribution: Information diffusion from template to search for visual object tracking publication-title: Int. J. Netw. Dyn. Intell. – year: 2022 ident: b7 article-title: Capsule network based contrastive learning of unsupervised visual representations – reference: Y. LeCun, F. Huang, L. Bottou, Learning methods for generic object recognition with invariance to pose and lighting, in: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, Vol. 2, 2004, p. 104. – volume: 12 year: 2024 ident: b20 article-title: MASNet: mixed attention Siamese network for visual object tracking publication-title: Syst. Sci. Control Eng. – year: 2014 ident: b36 article-title: Adam: A method for stochastic optimization – year: 2024 ident: b10 article-title: Capsule networks with residual pose routing publication-title: IEEE Trans. Neural Netw. Learn. Syst. – reference: H. Liu, C. Li, X. Zhang, et al., Depression Detection via Capsule Networks with Contrastive Learning, in: Proceedings of the AAAI Conference on Artificial Intelligence, Vol. 38, (20) 2024, pp. 22231–22239. – reference: J. Gu, V. Tresp, H. Hu, Capsule network is not more robust than convolutional network, in: IEEE Conference on Computer Vision and Pattern Recognition, 2021, pp. 14309–14317. – volume: 148 year: 2024 ident: b14 article-title: Global routing between capsules publication-title: Pattern Recognit. – reference: T. Chen, S. Kornblith, M. Norouzi, G.E. Hinton, A simple framework for contrastive learning of visual representations, in: International Conference on Machine Learning, 2020, pp. 1597–1607. – reference: S. Xie, R. Girshick, K. He, et al., Aggregated residual transformations for deep neural networks, in: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2017, pp. 1492–1500. – volume: 563 year: 2024 ident: b11 article-title: Deep unsupervised part-whole relational visual saliency publication-title: Neurocomputing – year: 2022 ident: 10.1016/j.patcog.2024.111059_b19 – volume: 238 year: 2024 ident: 10.1016/j.patcog.2024.111059_b29 article-title: A non-iterative capsule network with interdependent agreement routing publication-title: Expert Syst. Appl. doi: 10.1016/j.eswa.2023.122284 – ident: 10.1016/j.patcog.2024.111059_b38 doi: 10.1109/CVPR46437.2021.01408 – year: 2022 ident: 10.1016/j.patcog.2024.111059_b7 – year: 2017 ident: 10.1016/j.patcog.2024.111059_b33 – ident: 10.1016/j.patcog.2024.111059_b23 doi: 10.1109/CVPR42600.2020.00975 – ident: 10.1016/j.patcog.2024.111059_b5 – ident: 10.1016/j.patcog.2024.111059_b9 doi: 10.1109/ICCV.2019.00580 – ident: 10.1016/j.patcog.2024.111059_b3 doi: 10.1007/978-3-642-21735-7_6 – year: 2019 ident: 10.1016/j.patcog.2024.111059_b40 article-title: openTSNE: a modular Python library for t-SNE dimensionality reduction and embedding publication-title: BioRxiv – volume: 124 year: 2022 ident: 10.1016/j.patcog.2024.111059_b21 article-title: Sparse CapsNet with explicit regularizer publication-title: Pattern Recognit. doi: 10.1016/j.patcog.2021.108486 – year: 2024 ident: 10.1016/j.patcog.2024.111059_b10 article-title: Capsule networks with residual pose routing publication-title: IEEE Trans. Neural Netw. Learn. Syst. – year: 2017 ident: 10.1016/j.patcog.2024.111059_b25 – ident: 10.1016/j.patcog.2024.111059_b35 – ident: 10.1016/j.patcog.2024.111059_b27 doi: 10.1109/CVPR.2017.634 – ident: 10.1016/j.patcog.2024.111059_b18 – volume: 35 start-page: 413 issue: 3 year: 2023 ident: 10.1016/j.patcog.2024.111059_b2 article-title: How to represent part-whole hierarchies in a neural network publication-title: Neural Comput. doi: 10.1162/neco_a_01557 – volume: 11 start-page: 14634 issue: 1 year: 2021 ident: 10.1016/j.patcog.2024.111059_b28 article-title: Efficient-capsnet: Capsule network with self-attention routing publication-title: Sci. Rep. doi: 10.1038/s41598-021-93977-0 – year: 2019 ident: 10.1016/j.patcog.2024.111059_b31 – volume: 148 year: 2024 ident: 10.1016/j.patcog.2024.111059_b14 article-title: Global routing between capsules publication-title: Pattern Recognit. doi: 10.1016/j.patcog.2023.110142 – volume: 18 start-page: 4383 year: 2022 ident: 10.1016/j.patcog.2024.111059_b22 article-title: A fast routing capsule network with improved dense blocks publication-title: IEEE Trans. Ind. Inform. doi: 10.1109/TII.2021.3128412 – ident: 10.1016/j.patcog.2024.111059_b26 – start-page: 1 year: 2024 ident: 10.1016/j.patcog.2024.111059_b15 article-title: A robust interpolated model predictive control based on recurrent neural networks for a nonholonomic differential-drive mobile robot with quasi-LPV representation: computational complexity and conservatism publication-title: Int. J. Syst. Sci. – ident: 10.1016/j.patcog.2024.111059_b30 doi: 10.1109/CVPR.2004.1315150 – ident: 10.1016/j.patcog.2024.111059_b6 doi: 10.1609/aaai.v38i20.30228 – volume: 86 start-page: 2278 issue: 11 year: 1998 ident: 10.1016/j.patcog.2024.111059_b32 article-title: Gradient-based learning applied to document recognition publication-title: Proc. IEEE doi: 10.1109/5.726791 – volume: 137 year: 2023 ident: 10.1016/j.patcog.2024.111059_b12 article-title: Hybrid routing transformer for zero-shot learning publication-title: Pattern Recognit. doi: 10.1016/j.patcog.2022.109270 – volume: 32 start-page: 7658 year: 2019 ident: 10.1016/j.patcog.2024.111059_b37 article-title: Self-routing capsule networks publication-title: Adv. Neural Inf. Process. Syst. – volume: 3 issue: 1 year: 2024 ident: 10.1016/j.patcog.2024.111059_b16 article-title: Learning regression distribution: Information diffusion from template to search for visual object tracking publication-title: Int. J. Netw. Dyn. Intell. – year: 2014 ident: 10.1016/j.patcog.2024.111059_b36 – ident: 10.1016/j.patcog.2024.111059_b24 doi: 10.1109/CVPR.2016.90 – volume: 141 year: 2024 ident: 10.1016/j.patcog.2024.111059_b8 article-title: RPI-CapsuleGAN: Predicting RNA-protein interactions through an interpretable generative adversarial capsule network publication-title: Pattern Recognit. – year: 2011 ident: 10.1016/j.patcog.2024.111059_b39 – volume: 32 year: 2019 ident: 10.1016/j.patcog.2024.111059_b4 article-title: Stacked capsule autoencoders publication-title: Adv. Neural Inf. Process. Syst. – ident: 10.1016/j.patcog.2024.111059_b17 – ident: 10.1016/j.patcog.2024.111059_b34 – volume: 563 year: 2024 ident: 10.1016/j.patcog.2024.111059_b11 article-title: Deep unsupervised part-whole relational visual saliency publication-title: Neurocomputing doi: 10.1016/j.neucom.2023.126916 – volume: 12 issue: 1 year: 2024 ident: 10.1016/j.patcog.2024.111059_b20 article-title: MASNet: mixed attention Siamese network for visual object tracking publication-title: Syst. Sci. Control Eng. – volume: 34 start-page: 24993 year: 2021 ident: 10.1016/j.patcog.2024.111059_b1 article-title: Canonical capsules: Self-supervised capsules in canonical pose publication-title: Adv. Neural Inf. Process. Syst. – volume: 44 start-page: 3688 issue: 7 year: 2021 ident: 10.1016/j.patcog.2024.111059_b13 article-title: Part-object relational visual saliency publication-title: IEEE Trans. Pattern Anal. Mach. Intell. |
| SSID | ssj0017142 |
| Score | 2.4743958 |
| Snippet | Limited attention has been given to unsupervised capsule networks (CapsNets) with contrastive learning due to the challenge of harmoniously learning... |
| SourceID | crossref elsevier |
| SourceType | Enrichment Source Index Database Publisher |
| StartPage | 111059 |
| SubjectTerms | Comprehensive contrastive loss Interpretability Primary capsules Routing algorithm Unsupervised capsule network |
| Title | An interpretable unsupervised capsule network via comprehensive contrastive learning and two-stage training |
| URI | https://dx.doi.org/10.1016/j.patcog.2024.111059 |
| Volume | 158 |
| WOSCitedRecordID | wos001331367700001&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: 0031-3203 databaseCode: AIEXJ dateStart: 19950101 customDbUrl: isFulltext: true dateEnd: 99991231 titleUrlDefault: https://www.sciencedirect.com omitProxy: false ssIdentifier: ssj0017142 providerName: Elsevier |
| link | http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwtV3Nb9MwFLdKx4EL34jxJR-4IU-p7dT2sUJDwGGa0JDKLpHt2Fu7Kq3apuzP5yV2PsrQYAcuVvQUO1beL8_vOb_3jNB7qVJjrBkTpowjnLkR0ZJxQg2szsoljHtbHzYhTk7kdKpOBwPf5MLsFqIo5PW1Wv1XVYMMlF2lzt5B3e2gIIBrUDq0oHZo_0nxk6KuARGYhFVeVFlsylVlEjbgXFoNYTEIi0D__rCb6ZpWvnaXkcpek9f1pqYULZqNk5pn-XNJwJe8cO25En3P9rQu1Fklx0RGUvd__9wFg_KtbPdzIhH4R4fN88tONp-V_d0ImjYE5s7CshFhNGF7FjZUZ482EqxrEqqA3zDfYSdhfrSCZWh5AdE75Ufd7fvVsn9bxVpuYUNbm2dhlKwaJQuj3EMHVKRKDtHB5Mvx9Gv7v0mMeKgrH2ffJFnWTMCbs_mzE9NzTM4eo4cxosCTgIQnaOCKp-hRc1oHjsb7GbqaFHgPGLgPDByBgSMwMAAD7wED94CBG2BgAAZugYEbYDxH3z8dn338TOJRG8RCzLglzOeMKpdak8oqZkg8z7XwnCueeqWFAC86z5WiXiR6nFvOvJYgZ96OrBGGvUDDYlm4lwhz6scuN5waKbnUI5Mmnmqb04o5wKQ5RKx5d5mNdeir2S2y2zR3iEjbaxXqsPzlftGoJYu-ZPARM8DarT1f3fFJr9GD7kN4g4bbdeneovt2t51t1u8i0H4B5HWcwQ |
| 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=An+interpretable+unsupervised+capsule+network+via+comprehensive+contrastive+learning+and+two-stage+training&rft.jtitle=Pattern+recognition&rft.au=Zeng%2C+Ru&rft.au=Song%2C+Yan&rft.au=Zhong%2C+Yanjiu&rft.date=2025-02-01&rft.issn=0031-3203&rft.volume=158&rft.spage=111059&rft_id=info:doi/10.1016%2Fj.patcog.2024.111059&rft.externalDBID=n%2Fa&rft.externalDocID=10_1016_j_patcog_2024_111059 |
| thumbnail_l | http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/lc.gif&issn=0031-3203&client=summon |
| thumbnail_m | http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/mc.gif&issn=0031-3203&client=summon |
| thumbnail_s | http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/sc.gif&issn=0031-3203&client=summon |