Hierarchical few-shot learning based on coarse- and fine-grained relation network
Abstract Few-shot learning plays an important role in the field of machine learning. Many existing methods based on relation network achieve satisfactory results. However, these methods assume that classes are independent of each other and ignore their relationship. In this paper, we propose a hiera...
Ausführliche Beschreibung
Gespeichert in:
| Autor*in: |
Wu, Zhiping [verfasserIn] |
|---|
| Format: |
Artikel |
|---|---|
| Sprache: |
Englisch |
| Erschienen: |
2022 |
|---|
| Schlagwörter: |
|---|
| Anmerkung: |
© The Author(s), under exclusive licence to Springer Nature B.V. 2022 |
|---|
| Übergeordnetes Werk: |
Enthalten in: Artificial intelligence review - Springer Netherlands, 1987, 56(2022), 3 vom: 27. Juni, Seite 2011-2030 |
|---|---|
| Übergeordnetes Werk: |
volume:56 ; year:2022 ; number:3 ; day:27 ; month:06 ; pages:2011-2030 |
| Links: |
|---|
| DOI / URN: |
10.1007/s10462-022-10223-3 |
|---|
| Katalog-ID: |
OLC2134002816 |
|---|
| LEADER | 01000naa a22002652 4500 | ||
|---|---|---|---|
| 001 | OLC2134002816 | ||
| 003 | DE-627 | ||
| 005 | 20230506160113.0 | ||
| 007 | tu | ||
| 008 | 230506s2022 xx ||||| 00| ||eng c | ||
| 024 | 7 | |a 10.1007/s10462-022-10223-3 |2 doi | |
| 035 | |a (DE-627)OLC2134002816 | ||
| 035 | |a (DE-He213)s10462-022-10223-3-p | ||
| 040 | |a DE-627 |b ger |c DE-627 |e rakwb | ||
| 041 | |a eng | ||
| 082 | 0 | 4 | |a 004 |q VZ |
| 084 | |a 54.00 |2 bkl | ||
| 100 | 1 | |a Wu, Zhiping |e verfasserin |4 aut | |
| 245 | 1 | 0 | |a Hierarchical few-shot learning based on coarse- and fine-grained relation network |
| 264 | 1 | |c 2022 | |
| 336 | |a Text |b txt |2 rdacontent | ||
| 337 | |a ohne Hilfsmittel zu benutzen |b n |2 rdamedia | ||
| 338 | |a Band |b nc |2 rdacarrier | ||
| 500 | |a © The Author(s), under exclusive licence to Springer Nature B.V. 2022 | ||
| 520 | |a Abstract Few-shot learning plays an important role in the field of machine learning. Many existing methods based on relation network achieve satisfactory results. However, these methods assume that classes are independent of each other and ignore their relationship. In this paper, we propose a hierarchical few-shot learning model based on coarse- and fine-grained relation network (HCRN), which constructs a hierarchical structure by mining the relationship among different classes. Firstly, we extract deep and shallow features from different layers at a convolutional neural network. The shallow feature information contains more common features among similar classes, while the deep feature information is more specific. The complementary of these different types of data features can effectively construct coarse- and fine-grained structures by clustering. Secondly, we design coarse- and fine-grained relation networks to classify according to the guidance of the hierarchical structure. The hierarchical class structure learned from data is important auxiliary information for classification. Experimental results show that HCRN can outperform several state-of-the-art models on the Omniglot and miniImageNet datasets. Especially, HCRN obtains 6.47% improvement over the next best under the 5-way 1-shot setting on the miniImageNet dataset. | ||
| 650 | 4 | |a Few-shot learning | |
| 650 | 4 | |a Hierarchical structure | |
| 650 | 4 | |a Coarse- and fine-grained | |
| 650 | 4 | |a Relation network | |
| 700 | 1 | |a Zhao, Hong |0 (orcid)0000-0001-9339-1829 |4 aut | |
| 773 | 0 | 8 | |i Enthalten in |t Artificial intelligence review |d Springer Netherlands, 1987 |g 56(2022), 3 vom: 27. Juni, Seite 2011-2030 |w (DE-627)129223018 |w (DE-600)56633-0 |w (DE-576)014458209 |x 0269-2821 |7 nnns |
| 773 | 1 | 8 | |g volume:56 |g year:2022 |g number:3 |g day:27 |g month:06 |g pages:2011-2030 |
| 856 | 4 | 1 | |u https://doi.org/10.1007/s10462-022-10223-3 |z lizenzpflichtig |3 Volltext |
| 912 | |a GBV_USEFLAG_A | ||
| 912 | |a SYSFLAG_A | ||
| 912 | |a GBV_OLC | ||
| 912 | |a SSG-OLC-MAT | ||
| 936 | b | k | |a 54.00 |q VZ |
| 951 | |a AR | ||
| 952 | |d 56 |j 2022 |e 3 |b 27 |c 06 |h 2011-2030 | ||
| author_variant |
z w zw h z hz |
|---|---|
| matchkey_str |
article:02692821:2022----::irrhclesolannbsdnorenfnga |
| hierarchy_sort_str |
2022 |
| bklnumber |
54.00 |
| publishDate |
2022 |
| allfields |
10.1007/s10462-022-10223-3 doi (DE-627)OLC2134002816 (DE-He213)s10462-022-10223-3-p DE-627 ger DE-627 rakwb eng 004 VZ 54.00 bkl Wu, Zhiping verfasserin aut Hierarchical few-shot learning based on coarse- and fine-grained relation network 2022 Text txt rdacontent ohne Hilfsmittel zu benutzen n rdamedia Band nc rdacarrier © The Author(s), under exclusive licence to Springer Nature B.V. 2022 Abstract Few-shot learning plays an important role in the field of machine learning. Many existing methods based on relation network achieve satisfactory results. However, these methods assume that classes are independent of each other and ignore their relationship. In this paper, we propose a hierarchical few-shot learning model based on coarse- and fine-grained relation network (HCRN), which constructs a hierarchical structure by mining the relationship among different classes. Firstly, we extract deep and shallow features from different layers at a convolutional neural network. The shallow feature information contains more common features among similar classes, while the deep feature information is more specific. The complementary of these different types of data features can effectively construct coarse- and fine-grained structures by clustering. Secondly, we design coarse- and fine-grained relation networks to classify according to the guidance of the hierarchical structure. The hierarchical class structure learned from data is important auxiliary information for classification. Experimental results show that HCRN can outperform several state-of-the-art models on the Omniglot and miniImageNet datasets. Especially, HCRN obtains 6.47% improvement over the next best under the 5-way 1-shot setting on the miniImageNet dataset. Few-shot learning Hierarchical structure Coarse- and fine-grained Relation network Zhao, Hong (orcid)0000-0001-9339-1829 aut Enthalten in Artificial intelligence review Springer Netherlands, 1987 56(2022), 3 vom: 27. Juni, Seite 2011-2030 (DE-627)129223018 (DE-600)56633-0 (DE-576)014458209 0269-2821 nnns volume:56 year:2022 number:3 day:27 month:06 pages:2011-2030 https://doi.org/10.1007/s10462-022-10223-3 lizenzpflichtig Volltext GBV_USEFLAG_A SYSFLAG_A GBV_OLC SSG-OLC-MAT 54.00 VZ AR 56 2022 3 27 06 2011-2030 |
| spelling |
10.1007/s10462-022-10223-3 doi (DE-627)OLC2134002816 (DE-He213)s10462-022-10223-3-p DE-627 ger DE-627 rakwb eng 004 VZ 54.00 bkl Wu, Zhiping verfasserin aut Hierarchical few-shot learning based on coarse- and fine-grained relation network 2022 Text txt rdacontent ohne Hilfsmittel zu benutzen n rdamedia Band nc rdacarrier © The Author(s), under exclusive licence to Springer Nature B.V. 2022 Abstract Few-shot learning plays an important role in the field of machine learning. Many existing methods based on relation network achieve satisfactory results. However, these methods assume that classes are independent of each other and ignore their relationship. In this paper, we propose a hierarchical few-shot learning model based on coarse- and fine-grained relation network (HCRN), which constructs a hierarchical structure by mining the relationship among different classes. Firstly, we extract deep and shallow features from different layers at a convolutional neural network. The shallow feature information contains more common features among similar classes, while the deep feature information is more specific. The complementary of these different types of data features can effectively construct coarse- and fine-grained structures by clustering. Secondly, we design coarse- and fine-grained relation networks to classify according to the guidance of the hierarchical structure. The hierarchical class structure learned from data is important auxiliary information for classification. Experimental results show that HCRN can outperform several state-of-the-art models on the Omniglot and miniImageNet datasets. Especially, HCRN obtains 6.47% improvement over the next best under the 5-way 1-shot setting on the miniImageNet dataset. Few-shot learning Hierarchical structure Coarse- and fine-grained Relation network Zhao, Hong (orcid)0000-0001-9339-1829 aut Enthalten in Artificial intelligence review Springer Netherlands, 1987 56(2022), 3 vom: 27. Juni, Seite 2011-2030 (DE-627)129223018 (DE-600)56633-0 (DE-576)014458209 0269-2821 nnns volume:56 year:2022 number:3 day:27 month:06 pages:2011-2030 https://doi.org/10.1007/s10462-022-10223-3 lizenzpflichtig Volltext GBV_USEFLAG_A SYSFLAG_A GBV_OLC SSG-OLC-MAT 54.00 VZ AR 56 2022 3 27 06 2011-2030 |
| allfields_unstemmed |
10.1007/s10462-022-10223-3 doi (DE-627)OLC2134002816 (DE-He213)s10462-022-10223-3-p DE-627 ger DE-627 rakwb eng 004 VZ 54.00 bkl Wu, Zhiping verfasserin aut Hierarchical few-shot learning based on coarse- and fine-grained relation network 2022 Text txt rdacontent ohne Hilfsmittel zu benutzen n rdamedia Band nc rdacarrier © The Author(s), under exclusive licence to Springer Nature B.V. 2022 Abstract Few-shot learning plays an important role in the field of machine learning. Many existing methods based on relation network achieve satisfactory results. However, these methods assume that classes are independent of each other and ignore their relationship. In this paper, we propose a hierarchical few-shot learning model based on coarse- and fine-grained relation network (HCRN), which constructs a hierarchical structure by mining the relationship among different classes. Firstly, we extract deep and shallow features from different layers at a convolutional neural network. The shallow feature information contains more common features among similar classes, while the deep feature information is more specific. The complementary of these different types of data features can effectively construct coarse- and fine-grained structures by clustering. Secondly, we design coarse- and fine-grained relation networks to classify according to the guidance of the hierarchical structure. The hierarchical class structure learned from data is important auxiliary information for classification. Experimental results show that HCRN can outperform several state-of-the-art models on the Omniglot and miniImageNet datasets. Especially, HCRN obtains 6.47% improvement over the next best under the 5-way 1-shot setting on the miniImageNet dataset. Few-shot learning Hierarchical structure Coarse- and fine-grained Relation network Zhao, Hong (orcid)0000-0001-9339-1829 aut Enthalten in Artificial intelligence review Springer Netherlands, 1987 56(2022), 3 vom: 27. Juni, Seite 2011-2030 (DE-627)129223018 (DE-600)56633-0 (DE-576)014458209 0269-2821 nnns volume:56 year:2022 number:3 day:27 month:06 pages:2011-2030 https://doi.org/10.1007/s10462-022-10223-3 lizenzpflichtig Volltext GBV_USEFLAG_A SYSFLAG_A GBV_OLC SSG-OLC-MAT 54.00 VZ AR 56 2022 3 27 06 2011-2030 |
| allfieldsGer |
10.1007/s10462-022-10223-3 doi (DE-627)OLC2134002816 (DE-He213)s10462-022-10223-3-p DE-627 ger DE-627 rakwb eng 004 VZ 54.00 bkl Wu, Zhiping verfasserin aut Hierarchical few-shot learning based on coarse- and fine-grained relation network 2022 Text txt rdacontent ohne Hilfsmittel zu benutzen n rdamedia Band nc rdacarrier © The Author(s), under exclusive licence to Springer Nature B.V. 2022 Abstract Few-shot learning plays an important role in the field of machine learning. Many existing methods based on relation network achieve satisfactory results. However, these methods assume that classes are independent of each other and ignore their relationship. In this paper, we propose a hierarchical few-shot learning model based on coarse- and fine-grained relation network (HCRN), which constructs a hierarchical structure by mining the relationship among different classes. Firstly, we extract deep and shallow features from different layers at a convolutional neural network. The shallow feature information contains more common features among similar classes, while the deep feature information is more specific. The complementary of these different types of data features can effectively construct coarse- and fine-grained structures by clustering. Secondly, we design coarse- and fine-grained relation networks to classify according to the guidance of the hierarchical structure. The hierarchical class structure learned from data is important auxiliary information for classification. Experimental results show that HCRN can outperform several state-of-the-art models on the Omniglot and miniImageNet datasets. Especially, HCRN obtains 6.47% improvement over the next best under the 5-way 1-shot setting on the miniImageNet dataset. Few-shot learning Hierarchical structure Coarse- and fine-grained Relation network Zhao, Hong (orcid)0000-0001-9339-1829 aut Enthalten in Artificial intelligence review Springer Netherlands, 1987 56(2022), 3 vom: 27. Juni, Seite 2011-2030 (DE-627)129223018 (DE-600)56633-0 (DE-576)014458209 0269-2821 nnns volume:56 year:2022 number:3 day:27 month:06 pages:2011-2030 https://doi.org/10.1007/s10462-022-10223-3 lizenzpflichtig Volltext GBV_USEFLAG_A SYSFLAG_A GBV_OLC SSG-OLC-MAT 54.00 VZ AR 56 2022 3 27 06 2011-2030 |
| allfieldsSound |
10.1007/s10462-022-10223-3 doi (DE-627)OLC2134002816 (DE-He213)s10462-022-10223-3-p DE-627 ger DE-627 rakwb eng 004 VZ 54.00 bkl Wu, Zhiping verfasserin aut Hierarchical few-shot learning based on coarse- and fine-grained relation network 2022 Text txt rdacontent ohne Hilfsmittel zu benutzen n rdamedia Band nc rdacarrier © The Author(s), under exclusive licence to Springer Nature B.V. 2022 Abstract Few-shot learning plays an important role in the field of machine learning. Many existing methods based on relation network achieve satisfactory results. However, these methods assume that classes are independent of each other and ignore their relationship. In this paper, we propose a hierarchical few-shot learning model based on coarse- and fine-grained relation network (HCRN), which constructs a hierarchical structure by mining the relationship among different classes. Firstly, we extract deep and shallow features from different layers at a convolutional neural network. The shallow feature information contains more common features among similar classes, while the deep feature information is more specific. The complementary of these different types of data features can effectively construct coarse- and fine-grained structures by clustering. Secondly, we design coarse- and fine-grained relation networks to classify according to the guidance of the hierarchical structure. The hierarchical class structure learned from data is important auxiliary information for classification. Experimental results show that HCRN can outperform several state-of-the-art models on the Omniglot and miniImageNet datasets. Especially, HCRN obtains 6.47% improvement over the next best under the 5-way 1-shot setting on the miniImageNet dataset. Few-shot learning Hierarchical structure Coarse- and fine-grained Relation network Zhao, Hong (orcid)0000-0001-9339-1829 aut Enthalten in Artificial intelligence review Springer Netherlands, 1987 56(2022), 3 vom: 27. Juni, Seite 2011-2030 (DE-627)129223018 (DE-600)56633-0 (DE-576)014458209 0269-2821 nnns volume:56 year:2022 number:3 day:27 month:06 pages:2011-2030 https://doi.org/10.1007/s10462-022-10223-3 lizenzpflichtig Volltext GBV_USEFLAG_A SYSFLAG_A GBV_OLC SSG-OLC-MAT 54.00 VZ AR 56 2022 3 27 06 2011-2030 |
| language |
English |
| source |
Enthalten in Artificial intelligence review 56(2022), 3 vom: 27. Juni, Seite 2011-2030 volume:56 year:2022 number:3 day:27 month:06 pages:2011-2030 |
| sourceStr |
Enthalten in Artificial intelligence review 56(2022), 3 vom: 27. Juni, Seite 2011-2030 volume:56 year:2022 number:3 day:27 month:06 pages:2011-2030 |
| format_phy_str_mv |
Article |
| institution |
findex.gbv.de |
| topic_facet |
Few-shot learning Hierarchical structure Coarse- and fine-grained Relation network |
| dewey-raw |
004 |
| isfreeaccess_bool |
false |
| container_title |
Artificial intelligence review |
| authorswithroles_txt_mv |
Wu, Zhiping @@aut@@ Zhao, Hong @@aut@@ |
| publishDateDaySort_date |
2022-06-27T00:00:00Z |
| hierarchy_top_id |
129223018 |
| dewey-sort |
14 |
| id |
OLC2134002816 |
| language_de |
englisch |
| fullrecord |
<?xml version="1.0" encoding="UTF-8"?><collection xmlns="http://www.loc.gov/MARC21/slim"><record><leader>01000naa a22002652 4500</leader><controlfield tag="001">OLC2134002816</controlfield><controlfield tag="003">DE-627</controlfield><controlfield tag="005">20230506160113.0</controlfield><controlfield tag="007">tu</controlfield><controlfield tag="008">230506s2022 xx ||||| 00| ||eng c</controlfield><datafield tag="024" ind1="7" ind2=" "><subfield code="a">10.1007/s10462-022-10223-3</subfield><subfield code="2">doi</subfield></datafield><datafield tag="035" ind1=" " ind2=" "><subfield code="a">(DE-627)OLC2134002816</subfield></datafield><datafield tag="035" ind1=" " ind2=" "><subfield code="a">(DE-He213)s10462-022-10223-3-p</subfield></datafield><datafield tag="040" ind1=" " ind2=" "><subfield code="a">DE-627</subfield><subfield code="b">ger</subfield><subfield code="c">DE-627</subfield><subfield code="e">rakwb</subfield></datafield><datafield tag="041" ind1=" " ind2=" "><subfield code="a">eng</subfield></datafield><datafield tag="082" ind1="0" ind2="4"><subfield code="a">004</subfield><subfield code="q">VZ</subfield></datafield><datafield tag="084" ind1=" " ind2=" "><subfield code="a">54.00</subfield><subfield code="2">bkl</subfield></datafield><datafield tag="100" ind1="1" ind2=" "><subfield code="a">Wu, Zhiping</subfield><subfield code="e">verfasserin</subfield><subfield code="4">aut</subfield></datafield><datafield tag="245" ind1="1" ind2="0"><subfield code="a">Hierarchical few-shot learning based on coarse- and fine-grained relation network</subfield></datafield><datafield tag="264" ind1=" " ind2="1"><subfield code="c">2022</subfield></datafield><datafield tag="336" ind1=" " ind2=" "><subfield code="a">Text</subfield><subfield code="b">txt</subfield><subfield code="2">rdacontent</subfield></datafield><datafield tag="337" ind1=" " ind2=" "><subfield code="a">ohne Hilfsmittel zu benutzen</subfield><subfield code="b">n</subfield><subfield code="2">rdamedia</subfield></datafield><datafield tag="338" ind1=" " ind2=" "><subfield code="a">Band</subfield><subfield code="b">nc</subfield><subfield code="2">rdacarrier</subfield></datafield><datafield tag="500" ind1=" " ind2=" "><subfield code="a">© The Author(s), under exclusive licence to Springer Nature B.V. 2022</subfield></datafield><datafield tag="520" ind1=" " ind2=" "><subfield code="a">Abstract Few-shot learning plays an important role in the field of machine learning. Many existing methods based on relation network achieve satisfactory results. However, these methods assume that classes are independent of each other and ignore their relationship. In this paper, we propose a hierarchical few-shot learning model based on coarse- and fine-grained relation network (HCRN), which constructs a hierarchical structure by mining the relationship among different classes. Firstly, we extract deep and shallow features from different layers at a convolutional neural network. The shallow feature information contains more common features among similar classes, while the deep feature information is more specific. The complementary of these different types of data features can effectively construct coarse- and fine-grained structures by clustering. Secondly, we design coarse- and fine-grained relation networks to classify according to the guidance of the hierarchical structure. The hierarchical class structure learned from data is important auxiliary information for classification. Experimental results show that HCRN can outperform several state-of-the-art models on the Omniglot and miniImageNet datasets. Especially, HCRN obtains 6.47% improvement over the next best under the 5-way 1-shot setting on the miniImageNet dataset.</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Few-shot learning</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Hierarchical structure</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Coarse- and fine-grained</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Relation network</subfield></datafield><datafield tag="700" ind1="1" ind2=" "><subfield code="a">Zhao, Hong</subfield><subfield code="0">(orcid)0000-0001-9339-1829</subfield><subfield code="4">aut</subfield></datafield><datafield tag="773" ind1="0" ind2="8"><subfield code="i">Enthalten in</subfield><subfield code="t">Artificial intelligence review</subfield><subfield code="d">Springer Netherlands, 1987</subfield><subfield code="g">56(2022), 3 vom: 27. Juni, Seite 2011-2030</subfield><subfield code="w">(DE-627)129223018</subfield><subfield code="w">(DE-600)56633-0</subfield><subfield code="w">(DE-576)014458209</subfield><subfield code="x">0269-2821</subfield><subfield code="7">nnns</subfield></datafield><datafield tag="773" ind1="1" ind2="8"><subfield code="g">volume:56</subfield><subfield code="g">year:2022</subfield><subfield code="g">number:3</subfield><subfield code="g">day:27</subfield><subfield code="g">month:06</subfield><subfield code="g">pages:2011-2030</subfield></datafield><datafield tag="856" ind1="4" ind2="1"><subfield code="u">https://doi.org/10.1007/s10462-022-10223-3</subfield><subfield code="z">lizenzpflichtig</subfield><subfield code="3">Volltext</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_USEFLAG_A</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">SYSFLAG_A</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_OLC</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">SSG-OLC-MAT</subfield></datafield><datafield tag="936" ind1="b" ind2="k"><subfield code="a">54.00</subfield><subfield code="q">VZ</subfield></datafield><datafield tag="951" ind1=" " ind2=" "><subfield code="a">AR</subfield></datafield><datafield tag="952" ind1=" " ind2=" "><subfield code="d">56</subfield><subfield code="j">2022</subfield><subfield code="e">3</subfield><subfield code="b">27</subfield><subfield code="c">06</subfield><subfield code="h">2011-2030</subfield></datafield></record></collection>
|
| author |
Wu, Zhiping |
| spellingShingle |
Wu, Zhiping ddc 004 bkl 54.00 misc Few-shot learning misc Hierarchical structure misc Coarse- and fine-grained misc Relation network Hierarchical few-shot learning based on coarse- and fine-grained relation network |
| authorStr |
Wu, Zhiping |
| ppnlink_with_tag_str_mv |
@@773@@(DE-627)129223018 |
| format |
Article |
| dewey-ones |
004 - Data processing & computer science |
| delete_txt_mv |
keep |
| author_role |
aut aut |
| collection |
OLC |
| remote_str |
false |
| illustrated |
Not Illustrated |
| issn |
0269-2821 |
| topic_title |
004 VZ 54.00 bkl Hierarchical few-shot learning based on coarse- and fine-grained relation network Few-shot learning Hierarchical structure Coarse- and fine-grained Relation network |
| topic |
ddc 004 bkl 54.00 misc Few-shot learning misc Hierarchical structure misc Coarse- and fine-grained misc Relation network |
| topic_unstemmed |
ddc 004 bkl 54.00 misc Few-shot learning misc Hierarchical structure misc Coarse- and fine-grained misc Relation network |
| topic_browse |
ddc 004 bkl 54.00 misc Few-shot learning misc Hierarchical structure misc Coarse- and fine-grained misc Relation network |
| format_facet |
Aufsätze Gedruckte Aufsätze |
| format_main_str_mv |
Text Zeitschrift/Artikel |
| carriertype_str_mv |
nc |
| hierarchy_parent_title |
Artificial intelligence review |
| hierarchy_parent_id |
129223018 |
| dewey-tens |
000 - Computer science, knowledge & systems |
| hierarchy_top_title |
Artificial intelligence review |
| isfreeaccess_txt |
false |
| familylinks_str_mv |
(DE-627)129223018 (DE-600)56633-0 (DE-576)014458209 |
| title |
Hierarchical few-shot learning based on coarse- and fine-grained relation network |
| ctrlnum |
(DE-627)OLC2134002816 (DE-He213)s10462-022-10223-3-p |
| title_full |
Hierarchical few-shot learning based on coarse- and fine-grained relation network |
| author_sort |
Wu, Zhiping |
| journal |
Artificial intelligence review |
| journalStr |
Artificial intelligence review |
| lang_code |
eng |
| isOA_bool |
false |
| dewey-hundreds |
000 - Computer science, information & general works |
| recordtype |
marc |
| publishDateSort |
2022 |
| contenttype_str_mv |
txt |
| container_start_page |
2011 |
| author_browse |
Wu, Zhiping Zhao, Hong |
| container_volume |
56 |
| class |
004 VZ 54.00 bkl |
| format_se |
Aufsätze |
| author-letter |
Wu, Zhiping |
| doi_str_mv |
10.1007/s10462-022-10223-3 |
| normlink |
(ORCID)0000-0001-9339-1829 |
| normlink_prefix_str_mv |
(orcid)0000-0001-9339-1829 |
| dewey-full |
004 |
| title_sort |
hierarchical few-shot learning based on coarse- and fine-grained relation network |
| title_auth |
Hierarchical few-shot learning based on coarse- and fine-grained relation network |
| abstract |
Abstract Few-shot learning plays an important role in the field of machine learning. Many existing methods based on relation network achieve satisfactory results. However, these methods assume that classes are independent of each other and ignore their relationship. In this paper, we propose a hierarchical few-shot learning model based on coarse- and fine-grained relation network (HCRN), which constructs a hierarchical structure by mining the relationship among different classes. Firstly, we extract deep and shallow features from different layers at a convolutional neural network. The shallow feature information contains more common features among similar classes, while the deep feature information is more specific. The complementary of these different types of data features can effectively construct coarse- and fine-grained structures by clustering. Secondly, we design coarse- and fine-grained relation networks to classify according to the guidance of the hierarchical structure. The hierarchical class structure learned from data is important auxiliary information for classification. Experimental results show that HCRN can outperform several state-of-the-art models on the Omniglot and miniImageNet datasets. Especially, HCRN obtains 6.47% improvement over the next best under the 5-way 1-shot setting on the miniImageNet dataset. © The Author(s), under exclusive licence to Springer Nature B.V. 2022 |
| abstractGer |
Abstract Few-shot learning plays an important role in the field of machine learning. Many existing methods based on relation network achieve satisfactory results. However, these methods assume that classes are independent of each other and ignore their relationship. In this paper, we propose a hierarchical few-shot learning model based on coarse- and fine-grained relation network (HCRN), which constructs a hierarchical structure by mining the relationship among different classes. Firstly, we extract deep and shallow features from different layers at a convolutional neural network. The shallow feature information contains more common features among similar classes, while the deep feature information is more specific. The complementary of these different types of data features can effectively construct coarse- and fine-grained structures by clustering. Secondly, we design coarse- and fine-grained relation networks to classify according to the guidance of the hierarchical structure. The hierarchical class structure learned from data is important auxiliary information for classification. Experimental results show that HCRN can outperform several state-of-the-art models on the Omniglot and miniImageNet datasets. Especially, HCRN obtains 6.47% improvement over the next best under the 5-way 1-shot setting on the miniImageNet dataset. © The Author(s), under exclusive licence to Springer Nature B.V. 2022 |
| abstract_unstemmed |
Abstract Few-shot learning plays an important role in the field of machine learning. Many existing methods based on relation network achieve satisfactory results. However, these methods assume that classes are independent of each other and ignore their relationship. In this paper, we propose a hierarchical few-shot learning model based on coarse- and fine-grained relation network (HCRN), which constructs a hierarchical structure by mining the relationship among different classes. Firstly, we extract deep and shallow features from different layers at a convolutional neural network. The shallow feature information contains more common features among similar classes, while the deep feature information is more specific. The complementary of these different types of data features can effectively construct coarse- and fine-grained structures by clustering. Secondly, we design coarse- and fine-grained relation networks to classify according to the guidance of the hierarchical structure. The hierarchical class structure learned from data is important auxiliary information for classification. Experimental results show that HCRN can outperform several state-of-the-art models on the Omniglot and miniImageNet datasets. Especially, HCRN obtains 6.47% improvement over the next best under the 5-way 1-shot setting on the miniImageNet dataset. © The Author(s), under exclusive licence to Springer Nature B.V. 2022 |
| collection_details |
GBV_USEFLAG_A SYSFLAG_A GBV_OLC SSG-OLC-MAT |
| container_issue |
3 |
| title_short |
Hierarchical few-shot learning based on coarse- and fine-grained relation network |
| url |
https://doi.org/10.1007/s10462-022-10223-3 |
| remote_bool |
false |
| author2 |
Zhao, Hong |
| author2Str |
Zhao, Hong |
| ppnlink |
129223018 |
| mediatype_str_mv |
n |
| isOA_txt |
false |
| hochschulschrift_bool |
false |
| doi_str |
10.1007/s10462-022-10223-3 |
| up_date |
2024-07-03T22:52:27.493Z |
| _version_ |
1803600157731717120 |
| fullrecord_marcxml |
<?xml version="1.0" encoding="UTF-8"?><collection xmlns="http://www.loc.gov/MARC21/slim"><record><leader>01000naa a22002652 4500</leader><controlfield tag="001">OLC2134002816</controlfield><controlfield tag="003">DE-627</controlfield><controlfield tag="005">20230506160113.0</controlfield><controlfield tag="007">tu</controlfield><controlfield tag="008">230506s2022 xx ||||| 00| ||eng c</controlfield><datafield tag="024" ind1="7" ind2=" "><subfield code="a">10.1007/s10462-022-10223-3</subfield><subfield code="2">doi</subfield></datafield><datafield tag="035" ind1=" " ind2=" "><subfield code="a">(DE-627)OLC2134002816</subfield></datafield><datafield tag="035" ind1=" " ind2=" "><subfield code="a">(DE-He213)s10462-022-10223-3-p</subfield></datafield><datafield tag="040" ind1=" " ind2=" "><subfield code="a">DE-627</subfield><subfield code="b">ger</subfield><subfield code="c">DE-627</subfield><subfield code="e">rakwb</subfield></datafield><datafield tag="041" ind1=" " ind2=" "><subfield code="a">eng</subfield></datafield><datafield tag="082" ind1="0" ind2="4"><subfield code="a">004</subfield><subfield code="q">VZ</subfield></datafield><datafield tag="084" ind1=" " ind2=" "><subfield code="a">54.00</subfield><subfield code="2">bkl</subfield></datafield><datafield tag="100" ind1="1" ind2=" "><subfield code="a">Wu, Zhiping</subfield><subfield code="e">verfasserin</subfield><subfield code="4">aut</subfield></datafield><datafield tag="245" ind1="1" ind2="0"><subfield code="a">Hierarchical few-shot learning based on coarse- and fine-grained relation network</subfield></datafield><datafield tag="264" ind1=" " ind2="1"><subfield code="c">2022</subfield></datafield><datafield tag="336" ind1=" " ind2=" "><subfield code="a">Text</subfield><subfield code="b">txt</subfield><subfield code="2">rdacontent</subfield></datafield><datafield tag="337" ind1=" " ind2=" "><subfield code="a">ohne Hilfsmittel zu benutzen</subfield><subfield code="b">n</subfield><subfield code="2">rdamedia</subfield></datafield><datafield tag="338" ind1=" " ind2=" "><subfield code="a">Band</subfield><subfield code="b">nc</subfield><subfield code="2">rdacarrier</subfield></datafield><datafield tag="500" ind1=" " ind2=" "><subfield code="a">© The Author(s), under exclusive licence to Springer Nature B.V. 2022</subfield></datafield><datafield tag="520" ind1=" " ind2=" "><subfield code="a">Abstract Few-shot learning plays an important role in the field of machine learning. Many existing methods based on relation network achieve satisfactory results. However, these methods assume that classes are independent of each other and ignore their relationship. In this paper, we propose a hierarchical few-shot learning model based on coarse- and fine-grained relation network (HCRN), which constructs a hierarchical structure by mining the relationship among different classes. Firstly, we extract deep and shallow features from different layers at a convolutional neural network. The shallow feature information contains more common features among similar classes, while the deep feature information is more specific. The complementary of these different types of data features can effectively construct coarse- and fine-grained structures by clustering. Secondly, we design coarse- and fine-grained relation networks to classify according to the guidance of the hierarchical structure. The hierarchical class structure learned from data is important auxiliary information for classification. Experimental results show that HCRN can outperform several state-of-the-art models on the Omniglot and miniImageNet datasets. Especially, HCRN obtains 6.47% improvement over the next best under the 5-way 1-shot setting on the miniImageNet dataset.</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Few-shot learning</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Hierarchical structure</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Coarse- and fine-grained</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Relation network</subfield></datafield><datafield tag="700" ind1="1" ind2=" "><subfield code="a">Zhao, Hong</subfield><subfield code="0">(orcid)0000-0001-9339-1829</subfield><subfield code="4">aut</subfield></datafield><datafield tag="773" ind1="0" ind2="8"><subfield code="i">Enthalten in</subfield><subfield code="t">Artificial intelligence review</subfield><subfield code="d">Springer Netherlands, 1987</subfield><subfield code="g">56(2022), 3 vom: 27. Juni, Seite 2011-2030</subfield><subfield code="w">(DE-627)129223018</subfield><subfield code="w">(DE-600)56633-0</subfield><subfield code="w">(DE-576)014458209</subfield><subfield code="x">0269-2821</subfield><subfield code="7">nnns</subfield></datafield><datafield tag="773" ind1="1" ind2="8"><subfield code="g">volume:56</subfield><subfield code="g">year:2022</subfield><subfield code="g">number:3</subfield><subfield code="g">day:27</subfield><subfield code="g">month:06</subfield><subfield code="g">pages:2011-2030</subfield></datafield><datafield tag="856" ind1="4" ind2="1"><subfield code="u">https://doi.org/10.1007/s10462-022-10223-3</subfield><subfield code="z">lizenzpflichtig</subfield><subfield code="3">Volltext</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_USEFLAG_A</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">SYSFLAG_A</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_OLC</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">SSG-OLC-MAT</subfield></datafield><datafield tag="936" ind1="b" ind2="k"><subfield code="a">54.00</subfield><subfield code="q">VZ</subfield></datafield><datafield tag="951" ind1=" " ind2=" "><subfield code="a">AR</subfield></datafield><datafield tag="952" ind1=" " ind2=" "><subfield code="d">56</subfield><subfield code="j">2022</subfield><subfield code="e">3</subfield><subfield code="b">27</subfield><subfield code="c">06</subfield><subfield code="h">2011-2030</subfield></datafield></record></collection>
|
| score |
7.4004087 |