Temperature scaling unmixing framework based on convolutional autoencoder
•Different degrees of sparsity constraints are imposed adaptively in deep learning networks by temperature scaling techniques.•By considering the distribution of ground objects, it has good generalization in the real scene.•The framework is a new spatial level constraint method and can be transferre...
Gespeichert in:
| Veröffentlicht in: | International journal of applied earth observation and geoinformation Jg. 129; S. 103864 |
|---|---|
| Hauptverfasser: | , , , , |
| Format: | Journal Article |
| Sprache: | Englisch |
| Veröffentlicht: |
Elsevier B.V
01.05.2024
Elsevier |
| Schlagworte: | |
| ISSN: | 1569-8432, 1872-826X |
| Online-Zugang: | Volltext |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| Abstract | •Different degrees of sparsity constraints are imposed adaptively in deep learning networks by temperature scaling techniques.•By considering the distribution of ground objects, it has good generalization in the real scene.•The framework is a new spatial level constraint method and can be transferred to other convolutional autoencoder-based methods.
Hyperspectral unmixing is a key technology in the development of remote sensing applications. However, since both endmembers and abundances are unknown, unmixing is a non-convex problem with a large solution space. To solve this, existing methods usually impose the same strength of sparsity constraint. However, this often does not hold in practice. Because the abundances of purer regions are generally sparse, while the abundances distribution of more mixed regions should be smoother. Temperature scaling is a technique of introducing a temperature parameter T into softmax activation function to adjust the sparsity of the output. Inspired by this, we propose a temperature scaling unmixing (TSU) framework based on convolutional autoencoder (CAE). In this framework, sparse constraints of different intensities are applied to diverse regions by considering spatial similarity of ground objects distribution while preserving the ability of CAE to extract spatial features. What is more, equal-frequency binning is adopted to guide the division of regions by similarity matrix to realize the automatic temperature parameter setting. In addition, a CAE network is designed under the TSU framework in this paper, called TSUCAE. The TSUCAE method exhibits superior accuracy compared to state-of-the-art approaches, as demonstrated through extensive comparative experiments. Furthermore, the TSU framework can be transferred to other CAE-based unmixing methods directly while keeping the network structure of these methods unchanged. Sufficient ablation experiments also prove that the transfer of framework can improve the performance of unmixing. The code is publicly available at https://github.com/UPCGIT/TSUCAE. |
|---|---|
| AbstractList | •Different degrees of sparsity constraints are imposed adaptively in deep learning networks by temperature scaling techniques.•By considering the distribution of ground objects, it has good generalization in the real scene.•The framework is a new spatial level constraint method and can be transferred to other convolutional autoencoder-based methods.
Hyperspectral unmixing is a key technology in the development of remote sensing applications. However, since both endmembers and abundances are unknown, unmixing is a non-convex problem with a large solution space. To solve this, existing methods usually impose the same strength of sparsity constraint. However, this often does not hold in practice. Because the abundances of purer regions are generally sparse, while the abundances distribution of more mixed regions should be smoother. Temperature scaling is a technique of introducing a temperature parameter T into softmax activation function to adjust the sparsity of the output. Inspired by this, we propose a temperature scaling unmixing (TSU) framework based on convolutional autoencoder (CAE). In this framework, sparse constraints of different intensities are applied to diverse regions by considering spatial similarity of ground objects distribution while preserving the ability of CAE to extract spatial features. What is more, equal-frequency binning is adopted to guide the division of regions by similarity matrix to realize the automatic temperature parameter setting. In addition, a CAE network is designed under the TSU framework in this paper, called TSUCAE. The TSUCAE method exhibits superior accuracy compared to state-of-the-art approaches, as demonstrated through extensive comparative experiments. Furthermore, the TSU framework can be transferred to other CAE-based unmixing methods directly while keeping the network structure of these methods unchanged. Sufficient ablation experiments also prove that the transfer of framework can improve the performance of unmixing. The code is publicly available at https://github.com/UPCGIT/TSUCAE. Hyperspectral unmixing is a key technology in the development of remote sensing applications. However, since both endmembers and abundances are unknown, unmixing is a non-convex problem with a large solution space. To solve this, existing methods usually impose the same strength of sparsity constraint. However, this often does not hold in practice. Because the abundances of purer regions are generally sparse, while the abundances distribution of more mixed regions should be smoother. Temperature scaling is a technique of introducing a temperature parameter T into softmax activation function to adjust the sparsity of the output. Inspired by this, we propose a temperature scaling unmixing (TSU) framework based on convolutional autoencoder (CAE). In this framework, sparse constraints of different intensities are applied to diverse regions by considering spatial similarity of ground objects distribution while preserving the ability of CAE to extract spatial features. What is more, equal-frequency binning is adopted to guide the division of regions by similarity matrix to realize the automatic temperature parameter setting. In addition, a CAE network is designed under the TSU framework in this paper, called TSUCAE. The TSUCAE method exhibits superior accuracy compared to state-of-the-art approaches, as demonstrated through extensive comparative experiments. Furthermore, the TSU framework can be transferred to other CAE-based unmixing methods directly while keeping the network structure of these methods unchanged. Sufficient ablation experiments also prove that the transfer of framework can improve the performance of unmixing. The code is publicly available at https://github.com/UPCGIT/TSUCAE. |
| ArticleNumber | 103864 |
| Author | Sheng, Hui Yang, Zhiru Xu, Jin Xu, Mingming Liu, Shanwei |
| Author_xml | – sequence: 1 givenname: Jin surname: Xu fullname: Xu, Jin – sequence: 2 givenname: Mingming orcidid: 0000-0002-6758-9863 surname: Xu fullname: Xu, Mingming email: xumingming@upc.edu.cn, xumingming900405@126.com – sequence: 3 givenname: Shanwei surname: Liu fullname: Liu, Shanwei – sequence: 4 givenname: Hui surname: Sheng fullname: Sheng, Hui – sequence: 5 givenname: Zhiru surname: Yang fullname: Yang, Zhiru |
| BookMark | eNp9kUFv3CAQhVGVSE3S_IDefOzFG8DYYPVURW2yUqReUqk3NIbxCteGLeC0_fdl6yqHHHKBYTTfQ_PeJTnzwSMh7xndMcq6m2k3wWHHKRfl3ahOvCEXTEleK959Pyt12_W1Eg1_Sy5TmihlUnbqguwfcTlihLxGrJKB2flDtfrF_T4VY4QFf4X4oxogoa2Cr0zwT2Feswse5grWHNCbYDG-I-cjzAmv_99X5NuXz4-39_XD17v97aeH2oimz_Uw9ijtAFaIclKjhk4g78cGUQBTaC1TSow90E5aJbgUFAwVqqW8t8iH5orsN10bYNLH6BaIf3QAp_81QjxoiNmZGTWAbJtmkOMwCiGRgRpa1guOwijTjqJofdi0jjH8XDFlvbhkcJ7BY1iTbjjtOBXFqzLKtlETQ0oRx-evGdWnCPSkSwT6FIHeIiiMfMEYl-FkXY7g5lfJjxuJxcknh1En44rRaF1Ek8uq7hX6L5Nyo_0 |
| CitedBy_id | crossref_primary_10_1109_JSTARS_2025_3594155 crossref_primary_10_1109_TGRS_2025_3577325 crossref_primary_10_1007_s00024_025_03727_w |
| Cites_doi | 10.1109/ACCESS.2018.2818280 10.1109/TGRS.2020.2992743 10.1016/j.asoc.2018.05.012 10.1109/MSP.2022.3208987 10.1109/TGRS.2020.2977819 10.1109/TGRS.2019.2916296 10.1109/TPAMI.2005.165 10.1109/TGRS.2020.3015157 10.1109/JSTARS.2012.2192472 10.1109/JSTARS.2022.3175257 10.1109/JSTARS.2012.2194696 10.1109/LGRS.2017.2704625 10.1109/TGRS.2009.2014945 10.1109/TGRS.2011.2144605 10.1609/aaai.v37i12.26760 10.1016/j.ins.2013.03.014 10.1109/TGRS.2012.2213825 10.1109/MGRS.2022.3145854 10.1109/TSMC.1979.4310076 10.1109/LGRS.2019.2900733 10.1117/1.JRS.13.026509 10.1109/TGRS.2005.844293 10.1109/LGRS.2014.2325874 10.1109/JSTARS.2019.2901122 10.1109/JSTARS.2021.3140154 10.1109/TIP.2014.2363423 10.1109/TGRS.2015.2417162 10.1109/TGRS.2011.2162339 10.1117/12.157055 |
| ContentType | Journal Article |
| Copyright | 2024 |
| Copyright_xml | – notice: 2024 |
| DBID | 6I. AAFTH AAYXX CITATION 7S9 L.6 DOA |
| DOI | 10.1016/j.jag.2024.103864 |
| DatabaseName | ScienceDirect Open Access Titles Elsevier:ScienceDirect:Open Access CrossRef AGRICOLA AGRICOLA - Academic DOAJ Open Access Full Text |
| DatabaseTitle | CrossRef AGRICOLA AGRICOLA - Academic |
| DatabaseTitleList | AGRICOLA |
| Database_xml | – sequence: 1 dbid: DOA name: DOAJ Directory of Open Access Journals url: https://www.doaj.org/ sourceTypes: Open Website |
| DeliveryMethod | fulltext_linktorsrc |
| Discipline | Engineering Environmental Sciences |
| EISSN | 1872-826X |
| ExternalDocumentID | oai_doaj_org_article_aa7533b7fbf447e1a8b51942e4c8c5f4 10_1016_j_jag_2024_103864 S1569843224002188 |
| GroupedDBID | 0SF 29J 4.4 5GY 6I. AAFTH AAQXK AAXUO ABFYP ABLST ABQEM ABQYD ACLVX ACRLP ACSBN ADBBV ADMUD AFKWA AFTJW AFXIZ AGYEJ AHEUO AIKHN AJOXV AKIFW ALMA_UNASSIGNED_HOLDINGS AMFUW AMRAJ ASPBG ATOGT AVWKF AZFZN BKOJK BLECG EBS EJD FDB FEDTE FIRID FYGXN GROUPED_DOAJ HVGLF IMUCA KCYFY KOM M41 O-L P-8 P-9 P2P R2- RIG ROL SDF SDG SES SPC SSE SSJ T5K ~02 AAHBH AALRI AATTM AAXKI AAYWO AAYXX ABJNI ABWVN ACLOT ACRPL ADNMO ADVLN AEIPS AFJKZ AGQPQ AIIUN AITUG ANKPU APXCP CITATION EFJIC EFKBS 7S9 L.6 |
| ID | FETCH-LOGICAL-c439t-bf9e7dbad44dba0c8b64e29f3ee4a18edd1884f9a067d842740ac0485029de2b3 |
| IEDL.DBID | DOA |
| ISICitedReferencesCount | 4 |
| ISICitedReferencesURI | http://www.webofscience.com/api/gateway?GWVersion=2&SrcApp=Summon&SrcAuth=ProQuest&DestLinkType=CitingArticles&DestApp=WOS_CPL&KeyUT=001402753200001&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D |
| ISSN | 1569-8432 |
| IngestDate | Fri Oct 03 12:44:39 EDT 2025 Thu Oct 02 22:47:03 EDT 2025 Tue Nov 18 22:19:09 EST 2025 Sat Nov 29 03:32:47 EST 2025 Sat Jun 01 15:42:26 EDT 2024 |
| IsDoiOpenAccess | true |
| IsOpenAccess | true |
| IsPeerReviewed | true |
| IsScholarly | true |
| Keywords | Hyperspectral unmixing Temperature scaling Sparsity constraint Convolutional autoencoder Equal-frequency binning |
| Language | English |
| License | This is an open access article under the CC BY-NC-ND license. |
| LinkModel | DirectLink |
| MergedId | FETCHMERGED-LOGICAL-c439t-bf9e7dbad44dba0c8b64e29f3ee4a18edd1884f9a067d842740ac0485029de2b3 |
| Notes | ObjectType-Article-1 SourceType-Scholarly Journals-1 ObjectType-Feature-2 content type line 23 |
| ORCID | 0000-0002-6758-9863 |
| OpenAccessLink | https://doaj.org/article/aa7533b7fbf447e1a8b51942e4c8c5f4 |
| PQID | 3206204177 |
| PQPubID | 24069 |
| ParticipantIDs | doaj_primary_oai_doaj_org_article_aa7533b7fbf447e1a8b51942e4c8c5f4 proquest_miscellaneous_3206204177 crossref_primary_10_1016_j_jag_2024_103864 crossref_citationtrail_10_1016_j_jag_2024_103864 elsevier_sciencedirect_doi_10_1016_j_jag_2024_103864 |
| PublicationCentury | 2000 |
| PublicationDate | May 2024 2024-05-00 20240501 2024-05-01 |
| PublicationDateYYYYMMDD | 2024-05-01 |
| PublicationDate_xml | – month: 05 year: 2024 text: May 2024 |
| PublicationDecade | 2020 |
| PublicationTitle | International journal of applied earth observation and geoinformation |
| PublicationYear | 2024 |
| Publisher | Elsevier B.V Elsevier |
| Publisher_xml | – name: Elsevier B.V – name: Elsevier |
| References | Li, Li, Song, Mei, Du (b0095) 2019; 57 Rickard, L.J., Basedow, R.W., Zalewski, E.F., Silverglate, P.R., Landers, M., 1993. HYDICE: An airborne system for hyperspectral imaging. In: Proc SPIE Int Soc Opt Eng, Orlando, pp. 173–179. Martin, Plaza (b0110) 2012; 5 Palsson, Sigurdsson, Sveinsson, Ulfarsson (b0125) 2018; 6 Li, Agathos, Zaharie, Bioucas-Dias, Plaza, Li (b0090) 2015; 53 Palsson, Sveinsson, Ulfarsson (b0135) 2022; 15 Chen, Zhang, Zhang (b0015) 2023; 61 Shang, Huang, Shi, Liu, Liu, Steven, Sun, Xie, Qiao (b0160) 2023; 37 Zhu, Wang, Fan, Meng, Xiang, Pan (b0210) 2014; 23 Otsu (b0120) 1979; 9 Feng, Li, Wang, Du, Jia, Plaza (b0040) 2022; 15 Qian, Jia, Zhou, Robles-Kelly (b0140) 2011; 49 Zortea, Plaza (b0215) 2009; 47 Rajabi, Ghassemian (b0145) 2014; 12 Wu, Kao, Okuhara (b0180) 2013; 239 Dou, Gao, Zhang, Wang, Wang (b0035) 2020; 58 Maas, Hannun, Ng (b0105) 2013 Rao, Qu, Gao, Sun, Wu, Zhang (b0150) 2022; 106 Gao, Han, Hong, Zhang, Chanussot (b0045) 2022; 60 Zhang, Zhang, Tao, Huang (b0200) 2011; 50 Bioucas-Dias, Plaza, Dobigeon, Parente, Du, Gader, Chanussot (b0005) 2012; 5 Hong, Gao, Yao, Zhang, Plaza, Chanussot (b0065) 2020; 59 Ghosh, Roy, Koirala, Rasti, Scheunders (b0050) 2022; 60 Clark, Swayze, Wise, Livo, Hoefen, Kokaly, Sutley (b0025) 2003 Ioffe, Szegedy (b0070) 2015 Chen, Li, Ghamisi, Jia, Gu (b0010) 2017; 14 Lu, Wu, Yuan, Yan, Li (b0100) 2013; 51 Ismail, M., Ciesielski, V., 2003. An Empirical Investigation of the Impact of Discretization on Common Data Distributions. In: HIS, pp. 692–701. Khoshsokhan, Rajabi, Zayyani (b0085) 2019; 12 Yu, Ma, Mei, Fan, Huang, Li (b0190) 2022; 113 Palsson, Ulfarsson, Sveinsson (b0130) 2021; 59 Wang, Zhang, Feng (b0170) 2005; 27 Khoshsokhan, Rajabi, Zayyani (b0080) 2019; 13 He, Zhang, Ao, Huang (b0055) 2018; 70 Chen, Zhao, Wang, Richard, Rahardja (b0020) 2023; 40 Davis, Kavanaugh, Letelier, Bissett, Kohler (b0030) 2007 Tompson, Goroshin, Jain, LeCun, Bregler (b0165) 2015 Zhu, Jing, Bi (b0205) 2010 Xiang, Ali, Jung, Zhou (b0185) 2022; 60 Wang, Zhao, Chen, Rahardja (b0175) 2019; 16 Nascimento, Dias (b0115) 2005; 43 Zhang, Zhang (b0195) 2022; 10 Hinton, G., Vinyals, O., Dean, J., 2015. Distilling the Knowledge in a Neural Network. arXiv. Clark (10.1016/j.jag.2024.103864_b0025) 2003 Dou (10.1016/j.jag.2024.103864_b0035) 2020; 58 Zortea (10.1016/j.jag.2024.103864_b0215) 2009; 47 Wang (10.1016/j.jag.2024.103864_b0175) 2019; 16 Xiang (10.1016/j.jag.2024.103864_b0185) 2022; 60 Khoshsokhan (10.1016/j.jag.2024.103864_b0085) 2019; 12 Qian (10.1016/j.jag.2024.103864_b0140) 2011; 49 Tompson (10.1016/j.jag.2024.103864_b0165) 2015 Wu (10.1016/j.jag.2024.103864_b0180) 2013; 239 Yu (10.1016/j.jag.2024.103864_b0190) 2022; 113 Bioucas-Dias (10.1016/j.jag.2024.103864_b0005) 2012; 5 Ghosh (10.1016/j.jag.2024.103864_b0050) 2022; 60 Chen (10.1016/j.jag.2024.103864_b0015) 2023; 61 Zhu (10.1016/j.jag.2024.103864_b0205) 2010 Martin (10.1016/j.jag.2024.103864_b0110) 2012; 5 10.1016/j.jag.2024.103864_b0060 Gao (10.1016/j.jag.2024.103864_b0045) 2022; 60 He (10.1016/j.jag.2024.103864_b0055) 2018; 70 Rao (10.1016/j.jag.2024.103864_b0150) 2022; 106 Hong (10.1016/j.jag.2024.103864_b0065) 2020; 59 Otsu (10.1016/j.jag.2024.103864_b0120) 1979; 9 Wang (10.1016/j.jag.2024.103864_b0170) 2005; 27 10.1016/j.jag.2024.103864_b0155 Palsson (10.1016/j.jag.2024.103864_b0125) 2018; 6 Palsson (10.1016/j.jag.2024.103864_b0135) 2022; 15 10.1016/j.jag.2024.103864_b0075 Nascimento (10.1016/j.jag.2024.103864_b0115) 2005; 43 Lu (10.1016/j.jag.2024.103864_b0100) 2013; 51 Ioffe (10.1016/j.jag.2024.103864_b0070) 2015 Li (10.1016/j.jag.2024.103864_b0090) 2015; 53 Davis (10.1016/j.jag.2024.103864_b0030) 2007 Khoshsokhan (10.1016/j.jag.2024.103864_b0080) 2019; 13 Li (10.1016/j.jag.2024.103864_b0095) 2019; 57 Rajabi (10.1016/j.jag.2024.103864_b0145) 2014; 12 Zhang (10.1016/j.jag.2024.103864_b0200) 2011; 50 Zhang (10.1016/j.jag.2024.103864_b0195) 2022; 10 Zhu (10.1016/j.jag.2024.103864_b0210) 2014; 23 Palsson (10.1016/j.jag.2024.103864_b0130) 2021; 59 Feng (10.1016/j.jag.2024.103864_b0040) 2022; 15 Shang (10.1016/j.jag.2024.103864_b0160) 2023; 37 Chen (10.1016/j.jag.2024.103864_b0010) 2017; 14 Maas (10.1016/j.jag.2024.103864_b0105) 2013 Chen (10.1016/j.jag.2024.103864_b0020) 2023; 40 |
| References_xml | – volume: 12 start-page: 38 year: 2014 end-page: 42 ident: b0145 article-title: Spectral unmixing of hyperspectral imagery using multilayer NMF publication-title: IEEE Geosci. Remote Sens. Lett. – reference: Ismail, M., Ciesielski, V., 2003. An Empirical Investigation of the Impact of Discretization on Common Data Distributions. In: HIS, pp. 692–701. – volume: 15 start-page: 1340 year: 2022 end-page: 1372 ident: b0135 article-title: Blind hyperspectral unmixing using autoencoders: a critical comparison publication-title: IEEE J Sel. Top. Appl. Earth Obs. Remote Sens. – volume: 57 start-page: 7756 year: 2019 end-page: 7769 ident: b0095 article-title: Local spectral similarity preserving regularized robust sparse hyperspectral unmixing publication-title: IEEE Trans. Geosci. Remote Sens. – volume: 15 start-page: 4414 year: 2022 end-page: 4436 ident: b0040 article-title: Hyperspectral unmixing based on nonnegative matrix factorization: a comprehensive review publication-title: IEEE J Sel. Top. Appl. Earth Obs. Remote Sens. – volume: 50 start-page: 879 year: 2011 end-page: 893 ident: b0200 article-title: On combining multiple features for hyperspectral remote sensing image classification publication-title: IEEE Trans. Geosci. Remote Sens. – volume: 51 start-page: 2815 year: 2013 end-page: 2826 ident: b0100 article-title: Manifold regularized sparse NMF for hyperspectral unmixing publication-title: IEEE Trans. Geosci. Remote Sens. – volume: 5 start-page: 354 year: 2012 end-page: 379 ident: b0005 article-title: Hyperspectral unmixing overview: geometrical, statistical, and sparse regression-based approaches publication-title: IEEE J. Sel. Top. Appl. Earth Obs. Remote Sens. – volume: 37 start-page: 15082 year: 2023 end-page: 15090 ident: b0160 article-title: Improving training and inference of face recognition models via random temperature scaling publication-title: AAAI. – start-page: 648 year: 2015 end-page: 656 ident: b0165 article-title: Efficient object localization using convolutional networks publication-title: In: Proc IEEE Comput Soc Conf Comput Vision Pattern Recognit – volume: 60 start-page: 1 year: 2022 end-page: 18 ident: b0185 article-title: Hyperspectral anomaly detection with guided autoencoder publication-title: IEEE Trans. Geosci. Remote Sens. – volume: 23 start-page: 5412 year: 2014 end-page: 5427 ident: b0210 article-title: Spectral unmixing via data-guided sparsity publication-title: IEEE Trans. Image Process. – volume: 53 start-page: 5067 year: 2015 end-page: 5082 ident: b0090 article-title: Minimum volume simplex analysis: a fast algorithm for linear hyperspectral unmixing publication-title: IEEE Trans. Geosci. Remote Sens. – volume: 60 start-page: 1 year: 2022 end-page: 16 ident: b0050 article-title: Deep hyperspectral unmixing using transformer network publication-title: IEEE Trans. Geosci. Remote Sens. – start-page: 196 year: 2007 end-page: 207 ident: b0030 article-title: Spatial and spectral resolution considerations for imaging coastal waters publication-title: In: Proc SPIE Int Soc Opt Eng, San Diego – volume: 58 start-page: 6550 year: 2020 end-page: 6564 ident: b0035 article-title: Hyperspectral unmixing using orthogonal sparse prior-based autoencoder with hyper-laplacian loss and data-driven outlier detection publication-title: IEEE Trans. Geosci. Remote Sens. – volume: 10 start-page: 270 year: 2022 end-page: 294 ident: b0195 article-title: Artificial intelligence for remote sensing data analysis: a review of challenges and opportunities publication-title: IEEE Geosci. Remote Sens. Mag. – volume: 49 start-page: 4282 year: 2011 end-page: 4297 ident: b0140 article-title: Hyperspectral unmixing via L1,2 sparsity-constrained nonnegative matrix factorization publication-title: IEEE Trans. Geosci. Remote Sens. – volume: 113 year: 2022 ident: b0190 article-title: Multi-stage convolutional autoencoder network for hyperspectral unmixing publication-title: Int. J. Appl. Earth Obs. Geoinf. – volume: 59 start-page: 5966 year: 2020 end-page: 5978 ident: b0065 article-title: Graph convolutional networks for hyperspectral image classification publication-title: IEEE Trans. Geosci. Remote Sens. – volume: 47 start-page: 2679 year: 2009 end-page: 2693 ident: b0215 article-title: Spatial preprocessing for endmember extraction publication-title: IEEE Trans. Geosci. Remote Sens. – start-page: 448 year: 2015 end-page: 456 ident: b0070 article-title: Batch normalization: accelerating deep network training by reducing internal covariate shift publication-title: In: Int. Conf. Mach. Learn. – reference: Hinton, G., Vinyals, O., Dean, J., 2015. Distilling the Knowledge in a Neural Network. arXiv. – volume: 12 start-page: 1279 year: 2019 end-page: 1288 ident: b0085 article-title: Sparsity-constrained distributed unmixing of hyperspectral data publication-title: IEEE J Sel. Top. Appl. Earth Obs. Remote Sens. – volume: 40 start-page: 61 year: 2023 end-page: 74 ident: b0020 article-title: Integration of physics-based and data-driven models for hyperspectral image unmixing: a summary of current methods publication-title: IEEE Signal Process Mag. – volume: 6 start-page: 25646 year: 2018 end-page: 25656 ident: b0125 article-title: Hyperspectral unmixing using a neural network autoencoder publication-title: IEEE Access – volume: 16 start-page: 1467 year: 2019 end-page: 1471 ident: b0175 article-title: Nonlinear unmixing of hyperspectral data via deep autoencoder networks publication-title: IEEE Geosci. Remote Sens. Lett. – volume: 60 start-page: 1 year: 2022 end-page: 14 ident: b0045 article-title: CyCU-net: cycle-consistency unmixing network by learning cascaded autoencoders publication-title: IEEE Trans. Geosci. Remote Sens. – volume: 9 start-page: 62 year: 1979 end-page: 66 ident: b0120 article-title: A threshold selection method from gray-level histograms publication-title: IEEE Trans. Syst. Man Cybern: Syst. – volume: 106 year: 2022 ident: b0150 article-title: Transferable network with Siamese architecture for anomaly detection in hyperspectral images publication-title: Int. J. Appl. Earth Obs. Geoinf. – volume: 59 start-page: 535 year: 2021 end-page: 549 ident: b0130 article-title: Convolutional autoencoder for spectral-spatial hyperspectral unmixing publication-title: IEEE Trans. Geosci. Remote Sens. – volume: 27 start-page: 1334 year: 2005 end-page: 1339 ident: b0170 article-title: On the euclidean distance of images publication-title: IEEE Trans. Pattern Anal. Mach. Intell. – start-page: 3 year: 2013 ident: b0105 article-title: Rectifier nonlinearities improve neural network acoustic models publication-title: In: Proc. icml – start-page: 6183 year: 2010 end-page: 6188 ident: b0205 article-title: Exploration and improvement of Ostu threshold segmentation algorithm publication-title: In: Proc. World Congr. Intelligent Control Autom. WCICA – volume: 14 start-page: 1253 year: 2017 end-page: 1257 ident: b0010 article-title: Deep fusion of remote sensing data for accurate classification publication-title: IEEE Geosci. Remote Sens. Lett. – reference: Rickard, L.J., Basedow, R.W., Zalewski, E.F., Silverglate, P.R., Landers, M., 1993. HYDICE: An airborne system for hyperspectral imaging. In: Proc SPIE Int Soc Opt Eng, Orlando, pp. 173–179. – volume: 43 start-page: 898 year: 2005 end-page: 910 ident: b0115 article-title: Vertex component analysis: a fast algorithm to unmix hyperspectral data publication-title: IEEE Trans. Geosci. Remote Sens. – volume: 239 start-page: 154 year: 2013 end-page: 164 ident: b0180 article-title: Examination and comparison of conflicting data in granulated datasets: equal width interval vs. equal frequency interval publication-title: Inf. Sci. – start-page: 3395 year: 2003 ident: b0025 article-title: USGS digital spectral library splib06a – volume: 70 start-page: 80 year: 2018 end-page: 85 ident: b0055 article-title: Determining the optimal temperature parameter for Softmax function in reinforcement learning publication-title: Appl. Soft Comput. – volume: 5 start-page: 380 year: 2012 end-page: 395 ident: b0110 article-title: Spatial-spectral preprocessing prior to endmember identification and unmixing of remotely sensed hyperspectral data publication-title: IEEE J. Sel. Top. Appl. Earth Obs. Remote Sens. – volume: 61 start-page: 1 year: 2023 end-page: 14 ident: b0015 article-title: MSDformer: multiscale deformable transformer for hyperspectral image super-resolution publication-title: IEEE Trans. Geosci. Remote Sens. – volume: 13 start-page: 026509 year: 2019 ident: b0080 article-title: Clustered multitask nonnegative matrix factorization for spectral unmixing of hyperspectral data publication-title: J. Appl. Remote Sens. – volume: 6 start-page: 25646 year: 2018 ident: 10.1016/j.jag.2024.103864_b0125 article-title: Hyperspectral unmixing using a neural network autoencoder publication-title: IEEE Access doi: 10.1109/ACCESS.2018.2818280 – volume: 59 start-page: 535 issue: 1 year: 2021 ident: 10.1016/j.jag.2024.103864_b0130 article-title: Convolutional autoencoder for spectral-spatial hyperspectral unmixing publication-title: IEEE Trans. Geosci. Remote Sens. doi: 10.1109/TGRS.2020.2992743 – volume: 70 start-page: 80 year: 2018 ident: 10.1016/j.jag.2024.103864_b0055 article-title: Determining the optimal temperature parameter for Softmax function in reinforcement learning publication-title: Appl. Soft Comput. doi: 10.1016/j.asoc.2018.05.012 – volume: 40 start-page: 61 issue: 2 year: 2023 ident: 10.1016/j.jag.2024.103864_b0020 article-title: Integration of physics-based and data-driven models for hyperspectral image unmixing: a summary of current methods publication-title: IEEE Signal Process Mag. doi: 10.1109/MSP.2022.3208987 – volume: 58 start-page: 6550 issue: 9 year: 2020 ident: 10.1016/j.jag.2024.103864_b0035 article-title: Hyperspectral unmixing using orthogonal sparse prior-based autoencoder with hyper-laplacian loss and data-driven outlier detection publication-title: IEEE Trans. Geosci. Remote Sens. doi: 10.1109/TGRS.2020.2977819 – ident: 10.1016/j.jag.2024.103864_b0075 – volume: 57 start-page: 7756 issue: 10 year: 2019 ident: 10.1016/j.jag.2024.103864_b0095 article-title: Local spectral similarity preserving regularized robust sparse hyperspectral unmixing publication-title: IEEE Trans. Geosci. Remote Sens. doi: 10.1109/TGRS.2019.2916296 – volume: 27 start-page: 1334 issue: 8 year: 2005 ident: 10.1016/j.jag.2024.103864_b0170 article-title: On the euclidean distance of images publication-title: IEEE Trans. Pattern Anal. Mach. Intell. doi: 10.1109/TPAMI.2005.165 – volume: 60 start-page: 1 year: 2022 ident: 10.1016/j.jag.2024.103864_b0050 article-title: Deep hyperspectral unmixing using transformer network publication-title: IEEE Trans. Geosci. Remote Sens. – volume: 60 start-page: 1 year: 2022 ident: 10.1016/j.jag.2024.103864_b0045 article-title: CyCU-net: cycle-consistency unmixing network by learning cascaded autoencoders publication-title: IEEE Trans. Geosci. Remote Sens. – start-page: 3395 year: 2003 ident: 10.1016/j.jag.2024.103864_b0025 – volume: 59 start-page: 5966 issue: 7 year: 2020 ident: 10.1016/j.jag.2024.103864_b0065 article-title: Graph convolutional networks for hyperspectral image classification publication-title: IEEE Trans. Geosci. Remote Sens. doi: 10.1109/TGRS.2020.3015157 – volume: 5 start-page: 380 issue: 2 year: 2012 ident: 10.1016/j.jag.2024.103864_b0110 article-title: Spatial-spectral preprocessing prior to endmember identification and unmixing of remotely sensed hyperspectral data publication-title: IEEE J. Sel. Top. Appl. Earth Obs. Remote Sens. doi: 10.1109/JSTARS.2012.2192472 – start-page: 448 year: 2015 ident: 10.1016/j.jag.2024.103864_b0070 article-title: Batch normalization: accelerating deep network training by reducing internal covariate shift publication-title: In: Int. Conf. Mach. Learn. – volume: 15 start-page: 4414 year: 2022 ident: 10.1016/j.jag.2024.103864_b0040 article-title: Hyperspectral unmixing based on nonnegative matrix factorization: a comprehensive review publication-title: IEEE J Sel. Top. Appl. Earth Obs. Remote Sens. doi: 10.1109/JSTARS.2022.3175257 – volume: 5 start-page: 354 issue: 2 year: 2012 ident: 10.1016/j.jag.2024.103864_b0005 article-title: Hyperspectral unmixing overview: geometrical, statistical, and sparse regression-based approaches publication-title: IEEE J. Sel. Top. Appl. Earth Obs. Remote Sens. doi: 10.1109/JSTARS.2012.2194696 – start-page: 6183 year: 2010 ident: 10.1016/j.jag.2024.103864_b0205 article-title: Exploration and improvement of Ostu threshold segmentation algorithm publication-title: In: Proc. World Congr. Intelligent Control Autom. WCICA – volume: 14 start-page: 1253 issue: 8 year: 2017 ident: 10.1016/j.jag.2024.103864_b0010 article-title: Deep fusion of remote sensing data for accurate classification publication-title: IEEE Geosci. Remote Sens. Lett. doi: 10.1109/LGRS.2017.2704625 – start-page: 196 year: 2007 ident: 10.1016/j.jag.2024.103864_b0030 article-title: Spatial and spectral resolution considerations for imaging coastal waters publication-title: In: Proc SPIE Int Soc Opt Eng, San Diego – start-page: 648 year: 2015 ident: 10.1016/j.jag.2024.103864_b0165 article-title: Efficient object localization using convolutional networks – volume: 47 start-page: 2679 issue: 8 year: 2009 ident: 10.1016/j.jag.2024.103864_b0215 article-title: Spatial preprocessing for endmember extraction publication-title: IEEE Trans. Geosci. Remote Sens. doi: 10.1109/TGRS.2009.2014945 – ident: 10.1016/j.jag.2024.103864_b0060 – volume: 49 start-page: 4282 issue: 11 year: 2011 ident: 10.1016/j.jag.2024.103864_b0140 article-title: Hyperspectral unmixing via L1,2 sparsity-constrained nonnegative matrix factorization publication-title: IEEE Trans. Geosci. Remote Sens. doi: 10.1109/TGRS.2011.2144605 – volume: 106 year: 2022 ident: 10.1016/j.jag.2024.103864_b0150 article-title: Transferable network with Siamese architecture for anomaly detection in hyperspectral images publication-title: Int. J. Appl. Earth Obs. Geoinf. – volume: 37 start-page: 15082 issue: 12 year: 2023 ident: 10.1016/j.jag.2024.103864_b0160 article-title: Improving training and inference of face recognition models via random temperature scaling publication-title: AAAI. doi: 10.1609/aaai.v37i12.26760 – volume: 113 year: 2022 ident: 10.1016/j.jag.2024.103864_b0190 article-title: Multi-stage convolutional autoencoder network for hyperspectral unmixing publication-title: Int. J. Appl. Earth Obs. Geoinf. – start-page: 3 year: 2013 ident: 10.1016/j.jag.2024.103864_b0105 article-title: Rectifier nonlinearities improve neural network acoustic models publication-title: In: Proc. icml – volume: 239 start-page: 154 year: 2013 ident: 10.1016/j.jag.2024.103864_b0180 article-title: Examination and comparison of conflicting data in granulated datasets: equal width interval vs. equal frequency interval publication-title: Inf. Sci. doi: 10.1016/j.ins.2013.03.014 – volume: 51 start-page: 2815 year: 2013 ident: 10.1016/j.jag.2024.103864_b0100 article-title: Manifold regularized sparse NMF for hyperspectral unmixing publication-title: IEEE Trans. Geosci. Remote Sens. doi: 10.1109/TGRS.2012.2213825 – volume: 10 start-page: 270 issue: 2 year: 2022 ident: 10.1016/j.jag.2024.103864_b0195 article-title: Artificial intelligence for remote sensing data analysis: a review of challenges and opportunities publication-title: IEEE Geosci. Remote Sens. Mag. doi: 10.1109/MGRS.2022.3145854 – volume: 9 start-page: 62 issue: 1 year: 1979 ident: 10.1016/j.jag.2024.103864_b0120 article-title: A threshold selection method from gray-level histograms publication-title: IEEE Trans. Syst. Man Cybern: Syst. doi: 10.1109/TSMC.1979.4310076 – volume: 16 start-page: 1467 issue: 9 year: 2019 ident: 10.1016/j.jag.2024.103864_b0175 article-title: Nonlinear unmixing of hyperspectral data via deep autoencoder networks publication-title: IEEE Geosci. Remote Sens. Lett. doi: 10.1109/LGRS.2019.2900733 – volume: 13 start-page: 026509 year: 2019 ident: 10.1016/j.jag.2024.103864_b0080 article-title: Clustered multitask nonnegative matrix factorization for spectral unmixing of hyperspectral data publication-title: J. Appl. Remote Sens. doi: 10.1117/1.JRS.13.026509 – volume: 43 start-page: 898 issue: 4 year: 2005 ident: 10.1016/j.jag.2024.103864_b0115 article-title: Vertex component analysis: a fast algorithm to unmix hyperspectral data publication-title: IEEE Trans. Geosci. Remote Sens. doi: 10.1109/TGRS.2005.844293 – volume: 12 start-page: 38 issue: 1 year: 2014 ident: 10.1016/j.jag.2024.103864_b0145 article-title: Spectral unmixing of hyperspectral imagery using multilayer NMF publication-title: IEEE Geosci. Remote Sens. Lett. doi: 10.1109/LGRS.2014.2325874 – volume: 12 start-page: 1279 year: 2019 ident: 10.1016/j.jag.2024.103864_b0085 article-title: Sparsity-constrained distributed unmixing of hyperspectral data publication-title: IEEE J Sel. Top. Appl. Earth Obs. Remote Sens. doi: 10.1109/JSTARS.2019.2901122 – volume: 60 start-page: 1 year: 2022 ident: 10.1016/j.jag.2024.103864_b0185 article-title: Hyperspectral anomaly detection with guided autoencoder publication-title: IEEE Trans. Geosci. Remote Sens. – volume: 15 start-page: 1340 year: 2022 ident: 10.1016/j.jag.2024.103864_b0135 article-title: Blind hyperspectral unmixing using autoencoders: a critical comparison publication-title: IEEE J Sel. Top. Appl. Earth Obs. Remote Sens. doi: 10.1109/JSTARS.2021.3140154 – volume: 23 start-page: 5412 issue: 12 year: 2014 ident: 10.1016/j.jag.2024.103864_b0210 article-title: Spectral unmixing via data-guided sparsity publication-title: IEEE Trans. Image Process. doi: 10.1109/TIP.2014.2363423 – volume: 61 start-page: 1 year: 2023 ident: 10.1016/j.jag.2024.103864_b0015 article-title: MSDformer: multiscale deformable transformer for hyperspectral image super-resolution publication-title: IEEE Trans. Geosci. Remote Sens. – volume: 53 start-page: 5067 issue: 9 year: 2015 ident: 10.1016/j.jag.2024.103864_b0090 article-title: Minimum volume simplex analysis: a fast algorithm for linear hyperspectral unmixing publication-title: IEEE Trans. Geosci. Remote Sens. doi: 10.1109/TGRS.2015.2417162 – volume: 50 start-page: 879 issue: 3 year: 2011 ident: 10.1016/j.jag.2024.103864_b0200 article-title: On combining multiple features for hyperspectral remote sensing image classification publication-title: IEEE Trans. Geosci. Remote Sens. doi: 10.1109/TGRS.2011.2162339 – ident: 10.1016/j.jag.2024.103864_b0155 doi: 10.1117/12.157055 |
| SSID | ssj0017768 |
| Score | 2.3941562 |
| Snippet | •Different degrees of sparsity constraints are imposed adaptively in deep learning networks by temperature scaling techniques.•By considering the distribution... Hyperspectral unmixing is a key technology in the development of remote sensing applications. However, since both endmembers and abundances are unknown,... |
| SourceID | doaj proquest crossref elsevier |
| SourceType | Open Website Aggregation Database Enrichment Source Index Database Publisher |
| StartPage | 103864 |
| SubjectTerms | Convolutional autoencoder design Equal-frequency binning extracts Hyperspectral unmixing methodology remote sensing solutions Sparsity constraint spatial data temperature Temperature scaling |
| Title | Temperature scaling unmixing framework based on convolutional autoencoder |
| URI | https://dx.doi.org/10.1016/j.jag.2024.103864 https://www.proquest.com/docview/3206204177 https://doaj.org/article/aa7533b7fbf447e1a8b51942e4c8c5f4 |
| Volume | 129 |
| WOSCitedRecordID | wos001402753200001&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: PRVAON databaseName: DOAJ Directory of Open Access Journals customDbUrl: eissn: 1872-826X dateEnd: 99991231 omitProxy: false ssIdentifier: ssj0017768 issn: 1569-8432 databaseCode: DOA dateStart: 20200101 isFulltext: true titleUrlDefault: https://www.doaj.org/ providerName: Directory of Open Access Journals |
| link | http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwrV1LS8QwEA4iHvQgPnF9UcGTUOym001yVFH0Ih4UvIU8JrKiXdmH-POdNK2vg168lFKmaZiZdL4hk28YO-QEi4Ua-FyWQVKC4kRuQbgcrRDOYMCCh6bZhLi-lvf36uZLq69YE5bogZPijo0hQF1aEWwAENg30hLoAI7gpKtCwwRaCNUlU-3-gRDpEFw1ULmEknf7mU1l16N5oMSQQ0MOPoBvEakh7v8WmH78opu4c7HCllvAmJ2kia6yOazX2NIXGsE1tnn-eVqNRNvlOllnV7dIqDixJmcTsgaJZ7P6efgWb0JXl5XFUOazUZ3FGvTWF2kgM5uOIs-lx_EGu7s4vz27zNveCbkjiDHNbVAovDUegK6Fk3YAyFUoEcH0JXrflxKCMhStvATKTQvjaDVXBVceuS032Xw9qnGLZYIgE1mB92kk4D5ILytubUy2HFSoeqzo9KddSywe-1s86a6C7FGTynVUuU4q77Gjj1deEqvGb8Kn0SgfgpEQu3lAbqJbN9F_uUmPQWdS3WKLhBloqOFv3z7ozK9p3cXNFFPjaDbRJS8ilT952vZ_zG-HLcbPplLKXTY_Hc9wjy241-lwMt5vnPsdMs7-sQ |
| linkProvider | Directory of Open Access Journals |
| 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=Temperature+scaling+unmixing+framework+based+on+convolutional+autoencoder&rft.jtitle=International+journal+of+applied+earth+observation+and+geoinformation&rft.au=Xu%2C+Xiangmin&rft.au=Xu%2C+Mingming&rft.au=Liu%2C+Shanwei&rft.au=Sheng%2C+Hui&rft.date=2024-05-01&rft.issn=1569-8432&rft.volume=129+p.103864-&rft_id=info:doi/10.1016%2Fj.jag.2024.103864&rft.externalDBID=NO_FULL_TEXT |
| thumbnail_l | http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/lc.gif&issn=1569-8432&client=summon |
| thumbnail_m | http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/mc.gif&issn=1569-8432&client=summon |
| thumbnail_s | http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/sc.gif&issn=1569-8432&client=summon |