Review of Human Action Recognition Based on Deep Learning
Human action recognition is one of the important topics in video understanding. It is widely used in video surveillance, human-computer interaction, motion analysis, and video information retrieval. According to the chara-cteristics of the backbone network, this paper introduces the latest research...
Ausführliche Beschreibung
Autor*in: |
QIAN Huifang, YI Jianping, FU Yunhu [verfasserIn] |
---|
Format: |
E-Artikel |
---|---|
Sprache: |
Chinesisch |
Erschienen: |
2021 |
---|
Schlagwörter: |
2d convolutional neural network (2d cnn) 3d convolutional neural net-work (3d cnn) |
---|
Übergeordnetes Werk: |
In: Jisuanji kexue yu tansuo - Journal of Computer Engineering and Applications Beijing Co., Ltd., Science Press, 2021, 15(2021), 3, Seite 438-455 |
---|---|
Übergeordnetes Werk: |
volume:15 ; year:2021 ; number:3 ; pages:438-455 |
Links: |
---|
DOI / URN: |
10.3778/j.issn.1673-9418.2009095 |
---|
Katalog-ID: |
DOAJ04897398X |
---|
LEADER | 01000caa a22002652 4500 | ||
---|---|---|---|
001 | DOAJ04897398X | ||
003 | DE-627 | ||
005 | 20230308140911.0 | ||
007 | cr uuu---uuuuu | ||
008 | 230227s2021 xx |||||o 00| ||chi c | ||
024 | 7 | |a 10.3778/j.issn.1673-9418.2009095 |2 doi | |
035 | |a (DE-627)DOAJ04897398X | ||
035 | |a (DE-599)DOAJ9a08c6c7a69d4fa0b513e77815afafa9 | ||
040 | |a DE-627 |b ger |c DE-627 |e rakwb | ||
041 | |a chi | ||
050 | 0 | |a QA75.5-76.95 | |
100 | 0 | |a QIAN Huifang, YI Jianping, FU Yunhu |e verfasserin |4 aut | |
245 | 1 | 0 | |a Review of Human Action Recognition Based on Deep Learning |
264 | 1 | |c 2021 | |
336 | |a Text |b txt |2 rdacontent | ||
337 | |a Computermedien |b c |2 rdamedia | ||
338 | |a Online-Ressource |b cr |2 rdacarrier | ||
520 | |a Human action recognition is one of the important topics in video understanding. It is widely used in video surveillance, human-computer interaction, motion analysis, and video information retrieval. According to the chara-cteristics of the backbone network, this paper introduces the latest research results in the field of action recognition from three perspectives: 2D convolutional neural network, 3D convolutional neural network, and spatiotemporal decomposition network. And their advantages and disadvantages are qualitatively analyzed and compared. Then, from the two aspects of scene-related and temporal-related, the commonly used action video datasets are comprehensively summarized, and the characteristics and usage of different datasets are emphatically discussed. Subsequently, the common pre-training strategies in action recognition tasks are introduced, and the influence of pre-training techniques on the performance of action recognition models is emphatically analyzed. Finally, starting from the latest research trends, the future development direction of action recognition is discussed from six perspectives: fine-grained action recognition, streamlined model, few-shot learning, unsupervised learning, adaptive network, and video super-resolution action recognition. | ||
650 | 4 | |a human action recognition | |
650 | 4 | |a 2d convolutional neural network (2d cnn) | |
650 | 4 | |a 3d convolutional neural net-work (3d cnn) | |
650 | 4 | |a spatiotemporal decomposition network | |
650 | 4 | |a pre-training | |
653 | 0 | |a Electronic computers. Computer science | |
773 | 0 | 8 | |i In |t Jisuanji kexue yu tansuo |d Journal of Computer Engineering and Applications Beijing Co., Ltd., Science Press, 2021 |g 15(2021), 3, Seite 438-455 |w (DE-627)DOAJ078619211 |x 16739418 |7 nnns |
773 | 1 | 8 | |g volume:15 |g year:2021 |g number:3 |g pages:438-455 |
856 | 4 | 0 | |u https://doi.org/10.3778/j.issn.1673-9418.2009095 |z kostenfrei |
856 | 4 | 0 | |u https://doaj.org/article/9a08c6c7a69d4fa0b513e77815afafa9 |z kostenfrei |
856 | 4 | 0 | |u http://fcst.ceaj.org/CN/abstract/abstract2592.shtml |z kostenfrei |
856 | 4 | 2 | |u https://doaj.org/toc/1673-9418 |y Journal toc |z kostenfrei |
912 | |a GBV_USEFLAG_A | ||
912 | |a SYSFLAG_A | ||
912 | |a GBV_DOAJ | ||
951 | |a AR | ||
952 | |d 15 |j 2021 |e 3 |h 438-455 |
author_variant |
h y j f y q hyjfy hyjfyq |
---|---|
matchkey_str |
article:16739418:2021----::eiwfuaatorcgiinaeo |
hierarchy_sort_str |
2021 |
callnumber-subject-code |
QA |
publishDate |
2021 |
allfields |
10.3778/j.issn.1673-9418.2009095 doi (DE-627)DOAJ04897398X (DE-599)DOAJ9a08c6c7a69d4fa0b513e77815afafa9 DE-627 ger DE-627 rakwb chi QA75.5-76.95 QIAN Huifang, YI Jianping, FU Yunhu verfasserin aut Review of Human Action Recognition Based on Deep Learning 2021 Text txt rdacontent Computermedien c rdamedia Online-Ressource cr rdacarrier Human action recognition is one of the important topics in video understanding. It is widely used in video surveillance, human-computer interaction, motion analysis, and video information retrieval. According to the chara-cteristics of the backbone network, this paper introduces the latest research results in the field of action recognition from three perspectives: 2D convolutional neural network, 3D convolutional neural network, and spatiotemporal decomposition network. And their advantages and disadvantages are qualitatively analyzed and compared. Then, from the two aspects of scene-related and temporal-related, the commonly used action video datasets are comprehensively summarized, and the characteristics and usage of different datasets are emphatically discussed. Subsequently, the common pre-training strategies in action recognition tasks are introduced, and the influence of pre-training techniques on the performance of action recognition models is emphatically analyzed. Finally, starting from the latest research trends, the future development direction of action recognition is discussed from six perspectives: fine-grained action recognition, streamlined model, few-shot learning, unsupervised learning, adaptive network, and video super-resolution action recognition. human action recognition 2d convolutional neural network (2d cnn) 3d convolutional neural net-work (3d cnn) spatiotemporal decomposition network pre-training Electronic computers. Computer science In Jisuanji kexue yu tansuo Journal of Computer Engineering and Applications Beijing Co., Ltd., Science Press, 2021 15(2021), 3, Seite 438-455 (DE-627)DOAJ078619211 16739418 nnns volume:15 year:2021 number:3 pages:438-455 https://doi.org/10.3778/j.issn.1673-9418.2009095 kostenfrei https://doaj.org/article/9a08c6c7a69d4fa0b513e77815afafa9 kostenfrei http://fcst.ceaj.org/CN/abstract/abstract2592.shtml kostenfrei https://doaj.org/toc/1673-9418 Journal toc kostenfrei GBV_USEFLAG_A SYSFLAG_A GBV_DOAJ AR 15 2021 3 438-455 |
spelling |
10.3778/j.issn.1673-9418.2009095 doi (DE-627)DOAJ04897398X (DE-599)DOAJ9a08c6c7a69d4fa0b513e77815afafa9 DE-627 ger DE-627 rakwb chi QA75.5-76.95 QIAN Huifang, YI Jianping, FU Yunhu verfasserin aut Review of Human Action Recognition Based on Deep Learning 2021 Text txt rdacontent Computermedien c rdamedia Online-Ressource cr rdacarrier Human action recognition is one of the important topics in video understanding. It is widely used in video surveillance, human-computer interaction, motion analysis, and video information retrieval. According to the chara-cteristics of the backbone network, this paper introduces the latest research results in the field of action recognition from three perspectives: 2D convolutional neural network, 3D convolutional neural network, and spatiotemporal decomposition network. And their advantages and disadvantages are qualitatively analyzed and compared. Then, from the two aspects of scene-related and temporal-related, the commonly used action video datasets are comprehensively summarized, and the characteristics and usage of different datasets are emphatically discussed. Subsequently, the common pre-training strategies in action recognition tasks are introduced, and the influence of pre-training techniques on the performance of action recognition models is emphatically analyzed. Finally, starting from the latest research trends, the future development direction of action recognition is discussed from six perspectives: fine-grained action recognition, streamlined model, few-shot learning, unsupervised learning, adaptive network, and video super-resolution action recognition. human action recognition 2d convolutional neural network (2d cnn) 3d convolutional neural net-work (3d cnn) spatiotemporal decomposition network pre-training Electronic computers. Computer science In Jisuanji kexue yu tansuo Journal of Computer Engineering and Applications Beijing Co., Ltd., Science Press, 2021 15(2021), 3, Seite 438-455 (DE-627)DOAJ078619211 16739418 nnns volume:15 year:2021 number:3 pages:438-455 https://doi.org/10.3778/j.issn.1673-9418.2009095 kostenfrei https://doaj.org/article/9a08c6c7a69d4fa0b513e77815afafa9 kostenfrei http://fcst.ceaj.org/CN/abstract/abstract2592.shtml kostenfrei https://doaj.org/toc/1673-9418 Journal toc kostenfrei GBV_USEFLAG_A SYSFLAG_A GBV_DOAJ AR 15 2021 3 438-455 |
allfields_unstemmed |
10.3778/j.issn.1673-9418.2009095 doi (DE-627)DOAJ04897398X (DE-599)DOAJ9a08c6c7a69d4fa0b513e77815afafa9 DE-627 ger DE-627 rakwb chi QA75.5-76.95 QIAN Huifang, YI Jianping, FU Yunhu verfasserin aut Review of Human Action Recognition Based on Deep Learning 2021 Text txt rdacontent Computermedien c rdamedia Online-Ressource cr rdacarrier Human action recognition is one of the important topics in video understanding. It is widely used in video surveillance, human-computer interaction, motion analysis, and video information retrieval. According to the chara-cteristics of the backbone network, this paper introduces the latest research results in the field of action recognition from three perspectives: 2D convolutional neural network, 3D convolutional neural network, and spatiotemporal decomposition network. And their advantages and disadvantages are qualitatively analyzed and compared. Then, from the two aspects of scene-related and temporal-related, the commonly used action video datasets are comprehensively summarized, and the characteristics and usage of different datasets are emphatically discussed. Subsequently, the common pre-training strategies in action recognition tasks are introduced, and the influence of pre-training techniques on the performance of action recognition models is emphatically analyzed. Finally, starting from the latest research trends, the future development direction of action recognition is discussed from six perspectives: fine-grained action recognition, streamlined model, few-shot learning, unsupervised learning, adaptive network, and video super-resolution action recognition. human action recognition 2d convolutional neural network (2d cnn) 3d convolutional neural net-work (3d cnn) spatiotemporal decomposition network pre-training Electronic computers. Computer science In Jisuanji kexue yu tansuo Journal of Computer Engineering and Applications Beijing Co., Ltd., Science Press, 2021 15(2021), 3, Seite 438-455 (DE-627)DOAJ078619211 16739418 nnns volume:15 year:2021 number:3 pages:438-455 https://doi.org/10.3778/j.issn.1673-9418.2009095 kostenfrei https://doaj.org/article/9a08c6c7a69d4fa0b513e77815afafa9 kostenfrei http://fcst.ceaj.org/CN/abstract/abstract2592.shtml kostenfrei https://doaj.org/toc/1673-9418 Journal toc kostenfrei GBV_USEFLAG_A SYSFLAG_A GBV_DOAJ AR 15 2021 3 438-455 |
allfieldsGer |
10.3778/j.issn.1673-9418.2009095 doi (DE-627)DOAJ04897398X (DE-599)DOAJ9a08c6c7a69d4fa0b513e77815afafa9 DE-627 ger DE-627 rakwb chi QA75.5-76.95 QIAN Huifang, YI Jianping, FU Yunhu verfasserin aut Review of Human Action Recognition Based on Deep Learning 2021 Text txt rdacontent Computermedien c rdamedia Online-Ressource cr rdacarrier Human action recognition is one of the important topics in video understanding. It is widely used in video surveillance, human-computer interaction, motion analysis, and video information retrieval. According to the chara-cteristics of the backbone network, this paper introduces the latest research results in the field of action recognition from three perspectives: 2D convolutional neural network, 3D convolutional neural network, and spatiotemporal decomposition network. And their advantages and disadvantages are qualitatively analyzed and compared. Then, from the two aspects of scene-related and temporal-related, the commonly used action video datasets are comprehensively summarized, and the characteristics and usage of different datasets are emphatically discussed. Subsequently, the common pre-training strategies in action recognition tasks are introduced, and the influence of pre-training techniques on the performance of action recognition models is emphatically analyzed. Finally, starting from the latest research trends, the future development direction of action recognition is discussed from six perspectives: fine-grained action recognition, streamlined model, few-shot learning, unsupervised learning, adaptive network, and video super-resolution action recognition. human action recognition 2d convolutional neural network (2d cnn) 3d convolutional neural net-work (3d cnn) spatiotemporal decomposition network pre-training Electronic computers. Computer science In Jisuanji kexue yu tansuo Journal of Computer Engineering and Applications Beijing Co., Ltd., Science Press, 2021 15(2021), 3, Seite 438-455 (DE-627)DOAJ078619211 16739418 nnns volume:15 year:2021 number:3 pages:438-455 https://doi.org/10.3778/j.issn.1673-9418.2009095 kostenfrei https://doaj.org/article/9a08c6c7a69d4fa0b513e77815afafa9 kostenfrei http://fcst.ceaj.org/CN/abstract/abstract2592.shtml kostenfrei https://doaj.org/toc/1673-9418 Journal toc kostenfrei GBV_USEFLAG_A SYSFLAG_A GBV_DOAJ AR 15 2021 3 438-455 |
allfieldsSound |
10.3778/j.issn.1673-9418.2009095 doi (DE-627)DOAJ04897398X (DE-599)DOAJ9a08c6c7a69d4fa0b513e77815afafa9 DE-627 ger DE-627 rakwb chi QA75.5-76.95 QIAN Huifang, YI Jianping, FU Yunhu verfasserin aut Review of Human Action Recognition Based on Deep Learning 2021 Text txt rdacontent Computermedien c rdamedia Online-Ressource cr rdacarrier Human action recognition is one of the important topics in video understanding. It is widely used in video surveillance, human-computer interaction, motion analysis, and video information retrieval. According to the chara-cteristics of the backbone network, this paper introduces the latest research results in the field of action recognition from three perspectives: 2D convolutional neural network, 3D convolutional neural network, and spatiotemporal decomposition network. And their advantages and disadvantages are qualitatively analyzed and compared. Then, from the two aspects of scene-related and temporal-related, the commonly used action video datasets are comprehensively summarized, and the characteristics and usage of different datasets are emphatically discussed. Subsequently, the common pre-training strategies in action recognition tasks are introduced, and the influence of pre-training techniques on the performance of action recognition models is emphatically analyzed. Finally, starting from the latest research trends, the future development direction of action recognition is discussed from six perspectives: fine-grained action recognition, streamlined model, few-shot learning, unsupervised learning, adaptive network, and video super-resolution action recognition. human action recognition 2d convolutional neural network (2d cnn) 3d convolutional neural net-work (3d cnn) spatiotemporal decomposition network pre-training Electronic computers. Computer science In Jisuanji kexue yu tansuo Journal of Computer Engineering and Applications Beijing Co., Ltd., Science Press, 2021 15(2021), 3, Seite 438-455 (DE-627)DOAJ078619211 16739418 nnns volume:15 year:2021 number:3 pages:438-455 https://doi.org/10.3778/j.issn.1673-9418.2009095 kostenfrei https://doaj.org/article/9a08c6c7a69d4fa0b513e77815afafa9 kostenfrei http://fcst.ceaj.org/CN/abstract/abstract2592.shtml kostenfrei https://doaj.org/toc/1673-9418 Journal toc kostenfrei GBV_USEFLAG_A SYSFLAG_A GBV_DOAJ AR 15 2021 3 438-455 |
language |
Chinese |
source |
In Jisuanji kexue yu tansuo 15(2021), 3, Seite 438-455 volume:15 year:2021 number:3 pages:438-455 |
sourceStr |
In Jisuanji kexue yu tansuo 15(2021), 3, Seite 438-455 volume:15 year:2021 number:3 pages:438-455 |
format_phy_str_mv |
Article |
institution |
findex.gbv.de |
topic_facet |
human action recognition 2d convolutional neural network (2d cnn) 3d convolutional neural net-work (3d cnn) spatiotemporal decomposition network pre-training Electronic computers. Computer science |
isfreeaccess_bool |
true |
container_title |
Jisuanji kexue yu tansuo |
authorswithroles_txt_mv |
QIAN Huifang, YI Jianping, FU Yunhu @@aut@@ |
publishDateDaySort_date |
2021-01-01T00:00:00Z |
hierarchy_top_id |
DOAJ078619211 |
id |
DOAJ04897398X |
language_de |
chinesisch |
fullrecord |
<?xml version="1.0" encoding="UTF-8"?><collection xmlns="http://www.loc.gov/MARC21/slim"><record><leader>01000caa a22002652 4500</leader><controlfield tag="001">DOAJ04897398X</controlfield><controlfield tag="003">DE-627</controlfield><controlfield tag="005">20230308140911.0</controlfield><controlfield tag="007">cr uuu---uuuuu</controlfield><controlfield tag="008">230227s2021 xx |||||o 00| ||chi c</controlfield><datafield tag="024" ind1="7" ind2=" "><subfield code="a">10.3778/j.issn.1673-9418.2009095</subfield><subfield code="2">doi</subfield></datafield><datafield tag="035" ind1=" " ind2=" "><subfield code="a">(DE-627)DOAJ04897398X</subfield></datafield><datafield tag="035" ind1=" " ind2=" "><subfield code="a">(DE-599)DOAJ9a08c6c7a69d4fa0b513e77815afafa9</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">chi</subfield></datafield><datafield tag="050" ind1=" " ind2="0"><subfield code="a">QA75.5-76.95</subfield></datafield><datafield tag="100" ind1="0" ind2=" "><subfield code="a">QIAN Huifang, YI Jianping, FU Yunhu</subfield><subfield code="e">verfasserin</subfield><subfield code="4">aut</subfield></datafield><datafield tag="245" ind1="1" ind2="0"><subfield code="a">Review of Human Action Recognition Based on Deep Learning</subfield></datafield><datafield tag="264" ind1=" " ind2="1"><subfield code="c">2021</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">Computermedien</subfield><subfield code="b">c</subfield><subfield code="2">rdamedia</subfield></datafield><datafield tag="338" ind1=" " ind2=" "><subfield code="a">Online-Ressource</subfield><subfield code="b">cr</subfield><subfield code="2">rdacarrier</subfield></datafield><datafield tag="520" ind1=" " ind2=" "><subfield code="a">Human action recognition is one of the important topics in video understanding. It is widely used in video surveillance, human-computer interaction, motion analysis, and video information retrieval. According to the chara-cteristics of the backbone network, this paper introduces the latest research results in the field of action recognition from three perspectives: 2D convolutional neural network, 3D convolutional neural network, and spatiotemporal decomposition network. And their advantages and disadvantages are qualitatively analyzed and compared. Then, from the two aspects of scene-related and temporal-related, the commonly used action video datasets are comprehensively summarized, and the characteristics and usage of different datasets are emphatically discussed. Subsequently, the common pre-training strategies in action recognition tasks are introduced, and the influence of pre-training techniques on the performance of action recognition models is emphatically analyzed. Finally, starting from the latest research trends, the future development direction of action recognition is discussed from six perspectives: fine-grained action recognition, streamlined model, few-shot learning, unsupervised learning, adaptive network, and video super-resolution action recognition.</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">human action recognition</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">2d convolutional neural network (2d cnn)</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">3d convolutional neural net-work (3d cnn)</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">spatiotemporal decomposition network</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">pre-training</subfield></datafield><datafield tag="653" ind1=" " ind2="0"><subfield code="a">Electronic computers. Computer science</subfield></datafield><datafield tag="773" ind1="0" ind2="8"><subfield code="i">In</subfield><subfield code="t">Jisuanji kexue yu tansuo</subfield><subfield code="d">Journal of Computer Engineering and Applications Beijing Co., Ltd., Science Press, 2021</subfield><subfield code="g">15(2021), 3, Seite 438-455</subfield><subfield code="w">(DE-627)DOAJ078619211</subfield><subfield code="x">16739418</subfield><subfield code="7">nnns</subfield></datafield><datafield tag="773" ind1="1" ind2="8"><subfield code="g">volume:15</subfield><subfield code="g">year:2021</subfield><subfield code="g">number:3</subfield><subfield code="g">pages:438-455</subfield></datafield><datafield tag="856" ind1="4" ind2="0"><subfield code="u">https://doi.org/10.3778/j.issn.1673-9418.2009095</subfield><subfield code="z">kostenfrei</subfield></datafield><datafield tag="856" ind1="4" ind2="0"><subfield code="u">https://doaj.org/article/9a08c6c7a69d4fa0b513e77815afafa9</subfield><subfield code="z">kostenfrei</subfield></datafield><datafield tag="856" ind1="4" ind2="0"><subfield code="u">http://fcst.ceaj.org/CN/abstract/abstract2592.shtml</subfield><subfield code="z">kostenfrei</subfield></datafield><datafield tag="856" ind1="4" ind2="2"><subfield code="u">https://doaj.org/toc/1673-9418</subfield><subfield code="y">Journal toc</subfield><subfield code="z">kostenfrei</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_DOAJ</subfield></datafield><datafield tag="951" ind1=" " ind2=" "><subfield code="a">AR</subfield></datafield><datafield tag="952" ind1=" " ind2=" "><subfield code="d">15</subfield><subfield code="j">2021</subfield><subfield code="e">3</subfield><subfield code="h">438-455</subfield></datafield></record></collection>
|
callnumber-first |
Q - Science |
author |
QIAN Huifang, YI Jianping, FU Yunhu |
spellingShingle |
QIAN Huifang, YI Jianping, FU Yunhu misc QA75.5-76.95 misc human action recognition misc 2d convolutional neural network (2d cnn) misc 3d convolutional neural net-work (3d cnn) misc spatiotemporal decomposition network misc pre-training misc Electronic computers. Computer science Review of Human Action Recognition Based on Deep Learning |
authorStr |
QIAN Huifang, YI Jianping, FU Yunhu |
ppnlink_with_tag_str_mv |
@@773@@(DE-627)DOAJ078619211 |
format |
electronic Article |
delete_txt_mv |
keep |
author_role |
aut |
collection |
DOAJ |
remote_str |
true |
callnumber-label |
QA75 |
illustrated |
Not Illustrated |
issn |
16739418 |
topic_title |
QA75.5-76.95 Review of Human Action Recognition Based on Deep Learning human action recognition 2d convolutional neural network (2d cnn) 3d convolutional neural net-work (3d cnn) spatiotemporal decomposition network pre-training |
topic |
misc QA75.5-76.95 misc human action recognition misc 2d convolutional neural network (2d cnn) misc 3d convolutional neural net-work (3d cnn) misc spatiotemporal decomposition network misc pre-training misc Electronic computers. Computer science |
topic_unstemmed |
misc QA75.5-76.95 misc human action recognition misc 2d convolutional neural network (2d cnn) misc 3d convolutional neural net-work (3d cnn) misc spatiotemporal decomposition network misc pre-training misc Electronic computers. Computer science |
topic_browse |
misc QA75.5-76.95 misc human action recognition misc 2d convolutional neural network (2d cnn) misc 3d convolutional neural net-work (3d cnn) misc spatiotemporal decomposition network misc pre-training misc Electronic computers. Computer science |
format_facet |
Elektronische Aufsätze Aufsätze Elektronische Ressource |
format_main_str_mv |
Text Zeitschrift/Artikel |
carriertype_str_mv |
cr |
hierarchy_parent_title |
Jisuanji kexue yu tansuo |
hierarchy_parent_id |
DOAJ078619211 |
hierarchy_top_title |
Jisuanji kexue yu tansuo |
isfreeaccess_txt |
true |
familylinks_str_mv |
(DE-627)DOAJ078619211 |
title |
Review of Human Action Recognition Based on Deep Learning |
ctrlnum |
(DE-627)DOAJ04897398X (DE-599)DOAJ9a08c6c7a69d4fa0b513e77815afafa9 |
title_full |
Review of Human Action Recognition Based on Deep Learning |
author_sort |
QIAN Huifang, YI Jianping, FU Yunhu |
journal |
Jisuanji kexue yu tansuo |
journalStr |
Jisuanji kexue yu tansuo |
callnumber-first-code |
Q |
lang_code |
chi |
isOA_bool |
true |
recordtype |
marc |
publishDateSort |
2021 |
contenttype_str_mv |
txt |
container_start_page |
438 |
author_browse |
QIAN Huifang, YI Jianping, FU Yunhu |
container_volume |
15 |
class |
QA75.5-76.95 |
format_se |
Elektronische Aufsätze |
author-letter |
QIAN Huifang, YI Jianping, FU Yunhu |
doi_str_mv |
10.3778/j.issn.1673-9418.2009095 |
title_sort |
review of human action recognition based on deep learning |
callnumber |
QA75.5-76.95 |
title_auth |
Review of Human Action Recognition Based on Deep Learning |
abstract |
Human action recognition is one of the important topics in video understanding. It is widely used in video surveillance, human-computer interaction, motion analysis, and video information retrieval. According to the chara-cteristics of the backbone network, this paper introduces the latest research results in the field of action recognition from three perspectives: 2D convolutional neural network, 3D convolutional neural network, and spatiotemporal decomposition network. And their advantages and disadvantages are qualitatively analyzed and compared. Then, from the two aspects of scene-related and temporal-related, the commonly used action video datasets are comprehensively summarized, and the characteristics and usage of different datasets are emphatically discussed. Subsequently, the common pre-training strategies in action recognition tasks are introduced, and the influence of pre-training techniques on the performance of action recognition models is emphatically analyzed. Finally, starting from the latest research trends, the future development direction of action recognition is discussed from six perspectives: fine-grained action recognition, streamlined model, few-shot learning, unsupervised learning, adaptive network, and video super-resolution action recognition. |
abstractGer |
Human action recognition is one of the important topics in video understanding. It is widely used in video surveillance, human-computer interaction, motion analysis, and video information retrieval. According to the chara-cteristics of the backbone network, this paper introduces the latest research results in the field of action recognition from three perspectives: 2D convolutional neural network, 3D convolutional neural network, and spatiotemporal decomposition network. And their advantages and disadvantages are qualitatively analyzed and compared. Then, from the two aspects of scene-related and temporal-related, the commonly used action video datasets are comprehensively summarized, and the characteristics and usage of different datasets are emphatically discussed. Subsequently, the common pre-training strategies in action recognition tasks are introduced, and the influence of pre-training techniques on the performance of action recognition models is emphatically analyzed. Finally, starting from the latest research trends, the future development direction of action recognition is discussed from six perspectives: fine-grained action recognition, streamlined model, few-shot learning, unsupervised learning, adaptive network, and video super-resolution action recognition. |
abstract_unstemmed |
Human action recognition is one of the important topics in video understanding. It is widely used in video surveillance, human-computer interaction, motion analysis, and video information retrieval. According to the chara-cteristics of the backbone network, this paper introduces the latest research results in the field of action recognition from three perspectives: 2D convolutional neural network, 3D convolutional neural network, and spatiotemporal decomposition network. And their advantages and disadvantages are qualitatively analyzed and compared. Then, from the two aspects of scene-related and temporal-related, the commonly used action video datasets are comprehensively summarized, and the characteristics and usage of different datasets are emphatically discussed. Subsequently, the common pre-training strategies in action recognition tasks are introduced, and the influence of pre-training techniques on the performance of action recognition models is emphatically analyzed. Finally, starting from the latest research trends, the future development direction of action recognition is discussed from six perspectives: fine-grained action recognition, streamlined model, few-shot learning, unsupervised learning, adaptive network, and video super-resolution action recognition. |
collection_details |
GBV_USEFLAG_A SYSFLAG_A GBV_DOAJ |
container_issue |
3 |
title_short |
Review of Human Action Recognition Based on Deep Learning |
url |
https://doi.org/10.3778/j.issn.1673-9418.2009095 https://doaj.org/article/9a08c6c7a69d4fa0b513e77815afafa9 http://fcst.ceaj.org/CN/abstract/abstract2592.shtml https://doaj.org/toc/1673-9418 |
remote_bool |
true |
ppnlink |
DOAJ078619211 |
callnumber-subject |
QA - Mathematics |
mediatype_str_mv |
c |
isOA_txt |
true |
hochschulschrift_bool |
false |
doi_str |
10.3778/j.issn.1673-9418.2009095 |
callnumber-a |
QA75.5-76.95 |
up_date |
2024-07-03T20:41:38.247Z |
_version_ |
1803591927202840576 |
fullrecord_marcxml |
<?xml version="1.0" encoding="UTF-8"?><collection xmlns="http://www.loc.gov/MARC21/slim"><record><leader>01000caa a22002652 4500</leader><controlfield tag="001">DOAJ04897398X</controlfield><controlfield tag="003">DE-627</controlfield><controlfield tag="005">20230308140911.0</controlfield><controlfield tag="007">cr uuu---uuuuu</controlfield><controlfield tag="008">230227s2021 xx |||||o 00| ||chi c</controlfield><datafield tag="024" ind1="7" ind2=" "><subfield code="a">10.3778/j.issn.1673-9418.2009095</subfield><subfield code="2">doi</subfield></datafield><datafield tag="035" ind1=" " ind2=" "><subfield code="a">(DE-627)DOAJ04897398X</subfield></datafield><datafield tag="035" ind1=" " ind2=" "><subfield code="a">(DE-599)DOAJ9a08c6c7a69d4fa0b513e77815afafa9</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">chi</subfield></datafield><datafield tag="050" ind1=" " ind2="0"><subfield code="a">QA75.5-76.95</subfield></datafield><datafield tag="100" ind1="0" ind2=" "><subfield code="a">QIAN Huifang, YI Jianping, FU Yunhu</subfield><subfield code="e">verfasserin</subfield><subfield code="4">aut</subfield></datafield><datafield tag="245" ind1="1" ind2="0"><subfield code="a">Review of Human Action Recognition Based on Deep Learning</subfield></datafield><datafield tag="264" ind1=" " ind2="1"><subfield code="c">2021</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">Computermedien</subfield><subfield code="b">c</subfield><subfield code="2">rdamedia</subfield></datafield><datafield tag="338" ind1=" " ind2=" "><subfield code="a">Online-Ressource</subfield><subfield code="b">cr</subfield><subfield code="2">rdacarrier</subfield></datafield><datafield tag="520" ind1=" " ind2=" "><subfield code="a">Human action recognition is one of the important topics in video understanding. It is widely used in video surveillance, human-computer interaction, motion analysis, and video information retrieval. According to the chara-cteristics of the backbone network, this paper introduces the latest research results in the field of action recognition from three perspectives: 2D convolutional neural network, 3D convolutional neural network, and spatiotemporal decomposition network. And their advantages and disadvantages are qualitatively analyzed and compared. Then, from the two aspects of scene-related and temporal-related, the commonly used action video datasets are comprehensively summarized, and the characteristics and usage of different datasets are emphatically discussed. Subsequently, the common pre-training strategies in action recognition tasks are introduced, and the influence of pre-training techniques on the performance of action recognition models is emphatically analyzed. Finally, starting from the latest research trends, the future development direction of action recognition is discussed from six perspectives: fine-grained action recognition, streamlined model, few-shot learning, unsupervised learning, adaptive network, and video super-resolution action recognition.</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">human action recognition</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">2d convolutional neural network (2d cnn)</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">3d convolutional neural net-work (3d cnn)</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">spatiotemporal decomposition network</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">pre-training</subfield></datafield><datafield tag="653" ind1=" " ind2="0"><subfield code="a">Electronic computers. Computer science</subfield></datafield><datafield tag="773" ind1="0" ind2="8"><subfield code="i">In</subfield><subfield code="t">Jisuanji kexue yu tansuo</subfield><subfield code="d">Journal of Computer Engineering and Applications Beijing Co., Ltd., Science Press, 2021</subfield><subfield code="g">15(2021), 3, Seite 438-455</subfield><subfield code="w">(DE-627)DOAJ078619211</subfield><subfield code="x">16739418</subfield><subfield code="7">nnns</subfield></datafield><datafield tag="773" ind1="1" ind2="8"><subfield code="g">volume:15</subfield><subfield code="g">year:2021</subfield><subfield code="g">number:3</subfield><subfield code="g">pages:438-455</subfield></datafield><datafield tag="856" ind1="4" ind2="0"><subfield code="u">https://doi.org/10.3778/j.issn.1673-9418.2009095</subfield><subfield code="z">kostenfrei</subfield></datafield><datafield tag="856" ind1="4" ind2="0"><subfield code="u">https://doaj.org/article/9a08c6c7a69d4fa0b513e77815afafa9</subfield><subfield code="z">kostenfrei</subfield></datafield><datafield tag="856" ind1="4" ind2="0"><subfield code="u">http://fcst.ceaj.org/CN/abstract/abstract2592.shtml</subfield><subfield code="z">kostenfrei</subfield></datafield><datafield tag="856" ind1="4" ind2="2"><subfield code="u">https://doaj.org/toc/1673-9418</subfield><subfield code="y">Journal toc</subfield><subfield code="z">kostenfrei</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_DOAJ</subfield></datafield><datafield tag="951" ind1=" " ind2=" "><subfield code="a">AR</subfield></datafield><datafield tag="952" ind1=" " ind2=" "><subfield code="d">15</subfield><subfield code="j">2021</subfield><subfield code="e">3</subfield><subfield code="h">438-455</subfield></datafield></record></collection>
|
score |
7.3995714 |