DC-YOLOv3: A novel efficient object detection algorithm

Feature pyramids have become an essential component in most modern object detectors, such as Mask RCNN, YOLOv3, RetinaNet. In these detectors, the pyramidal feature representations are commonly used which represent an image with multi-scale feature layers. However, the detectors can’t be used in man...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:Journal of physics. Conference series Ročník 2082; číslo 1; s. 12012 - 12017
Hlavní autoři: Zhang, Xu, Han, Fang, Wang, Ping, Jiang, Wei, Wang, Chen
Médium: Journal Article
Jazyk:angličtina
Vydáno: IOP Publishing 01.11.2021
ISSN:1742-6588, 1742-6596
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Abstract Feature pyramids have become an essential component in most modern object detectors, such as Mask RCNN, YOLOv3, RetinaNet. In these detectors, the pyramidal feature representations are commonly used which represent an image with multi-scale feature layers. However, the detectors can’t be used in many real world applications which require real time performance under a computationally limited circumstance. In the paper, we study network architecture in YOLOv3 and modify the classical backbone--darknet53 of YOLOv3 by using a group of convolutions and dilated convolutions (DC). Then, a novel one-stage object detection network framework called DC-YOLOv3 is proposed. A lot of experiments on the Pascal 2017 benchmark prove the effectiveness of our framework. The results illustrate that DC-YOLOv3 achieves comparable results with YOLOv3 while being about 1.32× faster in training time and 1.38× faster in inference time.
AbstractList Feature pyramids have become an essential component in most modern object detectors, such as Mask RCNN, YOLOv3, RetinaNet. In these detectors, the pyramidal feature representations are commonly used which represent an image with multi-scale feature layers. However, the detectors can’t be used in many real world applications which require real time performance under a computationally limited circumstance. In the paper, we study network architecture in YOLOv3 and modify the classical backbone--darknet53 of YOLOv3 by using a group of convolutions and dilated convolutions (DC). Then, a novel one-stage object detection network framework called DC-YOLOv3 is proposed. A lot of experiments on the Pascal 2017 benchmark prove the effectiveness of our framework. The results illustrate that DC-YOLOv3 achieves comparable results with YOLOv3 while being about 1.32× faster in training time and 1.38× faster in inference time.
Author Wang, Chen
Han, Fang
Wang, Ping
Jiang, Wei
Zhang, Xu
Author_xml – sequence: 1
  givenname: Xu
  surname: Zhang
  fullname: Zhang, Xu
  organization: Science and education center of big data and computer application, Huanghe University of Science and Technology , China
– sequence: 2
  givenname: Fang
  surname: Han
  fullname: Han, Fang
  organization: Science and education center of big data and computer application, Huanghe University of Science and Technology , China
– sequence: 3
  givenname: Ping
  surname: Wang
  fullname: Wang, Ping
  organization: Science and education center of big data and computer application, Huanghe University of Science and Technology , China
– sequence: 4
  givenname: Wei
  surname: Jiang
  fullname: Jiang, Wei
  organization: Science and education center of big data and computer application, Huanghe University of Science and Technology , China
– sequence: 5
  givenname: Chen
  surname: Wang
  fullname: Wang, Chen
  organization: Science and education center of big data and computer application, Huanghe University of Science and Technology , China
BookMark eNqNUN9LwzAQDjLBbfo32GehNmnSJBV8GHX-YjBBffApJGmiGV0z0jrwvzdlMlAEPQ7u4L7v7r5vAkatbw0ApwieI8h5hhjJU1qUNMshzzOUQZTHPADj_WS07zk_ApOuW0GIY7AxYFdV-rJcLLf4Ipklrd-aJjHWOu1M2yderYzuk9r0sTjfJrJ59cH1b-tjcGhl05mTrzoFz9fzp-o2XSxv7qrZItUYwnjR1ooobVlJZK3yUhGMiDGqlpArXFKsGbdUclIwgihGOUWF5gZpKakpbIGn4HK3VwffdcFYoV0vh1_6IF0jEBSDC2LQJwatYnBBILFzIfLZD_4muLUMH_9gnu2Yzm_Eyr-HNuoU9w_V43eg2NQ2gvEv4L9OfAIY6n2t
CitedBy_id crossref_primary_10_3390_su16010117
Cites_doi 10.1007/s11263-013-0620-5
10.1109/TPAMI.2011.231
10.1007/s11263-014-0733-5
ContentType Journal Article
Copyright Published under licence by IOP Publishing Ltd
Copyright_xml – notice: Published under licence by IOP Publishing Ltd
DBID O3W
TSCCA
AAYXX
CITATION
DOI 10.1088/1742-6596/2082/1/012012
DatabaseName Institute of Physics Open Access Journal Titles
IOPscience (Open Access)
CrossRef
DatabaseTitle CrossRef
DatabaseTitleList CrossRef
Database_xml – sequence: 1
  dbid: O3W
  name: Institute of Physics Open Access Journal Titles
  url: http://iopscience.iop.org/
  sourceTypes:
    Enrichment Source
    Publisher
DeliveryMethod fulltext_linktorsrc
Discipline Physics
EISSN 1742-6596
ExternalDocumentID 10_1088_1742_6596_2082_1_012012
JPCS_2082_1_012012
GroupedDBID 1JI
29L
2WC
4.4
5B3
5GY
5PX
5VS
7.Q
AAJIO
AAJKP
ABHWH
ACAFW
ACHIP
AEFHF
AEJGL
AFKRA
AFYNE
AIYBF
AKPSB
ALMA_UNASSIGNED_HOLDINGS
ARAPS
ASPBG
ATQHT
AVWKF
AZFZN
BENPR
BGLVJ
CCPQU
CEBXE
CJUJL
CRLBU
CS3
DU5
E3Z
EBS
EDWGO
EQZZN
F5P
FRP
GROUPED_DOAJ
GX1
HCIFZ
HH5
IJHAN
IOP
IZVLO
J9A
KNG
KQ8
LAP
N5L
N9A
O3W
OK1
P2P
PIMPY
PJBAE
RIN
RNS
RO9
SY9
T37
TR2
TSCCA
UCJ
W28
XSB
~02
AAYXX
AEINN
AFFHD
CITATION
OVT
PHGZM
PHGZT
PQGLB
ROL
ID FETCH-LOGICAL-c3002-6fdb4bcf794adb29b4314eebda08b3963c78f6a8457416312615c8e1caa6e5f53
IEDL.DBID O3W
ISSN 1742-6588
IngestDate Tue Nov 18 22:09:56 EST 2025
Sat Nov 29 02:06:57 EST 2025
Wed Aug 21 03:43:00 EDT 2024
Fri Nov 12 21:34:55 EST 2021
IsDoiOpenAccess true
IsOpenAccess true
IsPeerReviewed true
IsScholarly true
Issue 1
Language English
License Content from this work may be used under the terms of the Creative Commons Attribution 3.0 licence. Any further distribution of this work must maintain attribution to the author(s) and the title of the work, journal citation and DOI.
LinkModel DirectLink
MergedId FETCHMERGED-LOGICAL-c3002-6fdb4bcf794adb29b4314eebda08b3963c78f6a8457416312615c8e1caa6e5f53
OpenAccessLink https://iopscience.iop.org/article/10.1088/1742-6596/2082/1/012012
PageCount 6
ParticipantIDs crossref_citationtrail_10_1088_1742_6596_2082_1_012012
crossref_primary_10_1088_1742_6596_2082_1_012012
iop_journals_10_1088_1742_6596_2082_1_012012
PublicationCentury 2000
PublicationDate 20211101
PublicationDateYYYYMMDD 2021-11-01
PublicationDate_xml – month: 11
  year: 2021
  text: 20211101
  day: 01
PublicationDecade 2020
PublicationTitle Journal of physics. Conference series
PublicationTitleAlternate J. Phys.: Conf. Ser
PublicationYear 2021
Publisher IOP Publishing
Publisher_xml – name: IOP Publishing
References Everingham (JPCS_2082_1_012012bib33) 2015; 111
Redmon (JPCS_2082_1_012012bib11) 2016
Redmon (JPCS_2082_1_012012bib28)
Redmon (JPCS_2082_1_012012bib13) 2018
Liu (JPCS_2082_1_012012bib15) 2016
Krizhevsky (JPCS_2082_1_012012bib2) 2012
Chen (JPCS_2082_1_012012bib32) 2016
Yu (JPCS_2082_1_012012bib31) 2015
Law (JPCS_2082_1_012012bib19) 2018
Shen (JPCS_2082_1_012012bib17) 2017
Yuan (JPCS_2082_1_012012bib26) 2019; 105
Lin (JPCS_2082_1_012012bib18) 2017
Zhao (JPCS_2082_1_012012bib16) 2019; 33
Bochkovskiy (JPCS_2082_1_012012bib29) 2020
Lin (JPCS_2082_1_012012bib20) 2017
Girshick (JPCS_2082_1_012012bib3) 2015
Liu (JPCS_2082_1_012012bib22) 2018
Kong (JPCS_2082_1_012012bib25) 2018
Ren (JPCS_2082_1_012012bib4) 2015
Dai (JPCS_2082_1_012012bib5) 2016
Cai (JPCS_2082_1_012012bib10) 2018
Zhang (JPCS_2082_1_012012bib8) 2016
Wang (JPCS_2082_1_012012bib30) 2018
Chen (JPCS_2082_1_012012bib21) 2021
Redmon (JPCS_2082_1_012012bib12) 2017
Yu (JPCS_2082_1_012012bib27) 2016
Zhao (JPCS_2082_1_012012bib23) 2019; 33
Carreira (JPCS_2082_1_012012bib7) 2012; 34
Uijlings (JPCS_2082_1_012012bib6) 2013; 104
He (JPCS_2082_1_012012bib9) 2017
Huang (JPCS_2082_1_012012bib14) 2018
Chen (JPCS_2082_1_012012bib24) 2018
Girshick (JPCS_2082_1_012012bib1) 2014
Lin (JPCS_2082_1_012012bib34) 2014
References_xml – start-page: 1440
  year: 2015
  ident: JPCS_2082_1_012012bib3
– volume: 104
  start-page: 154
  year: 2013
  ident: JPCS_2082_1_012012bib6
  article-title: Selective search for object recognition
  publication-title: International Journal of Computer Vision
  doi: 10.1007/s11263-013-0620-5
– year: 2016
  ident: JPCS_2082_1_012012bib15
– start-page: 734
  year: 2018
  ident: JPCS_2082_1_012012bib19
– year: 2018
  ident: JPCS_2082_1_012012bib24
– start-page: 1, 2, 3, 8
  year: 2017
  ident: JPCS_2082_1_012012bib18
– year: 2018
  ident: JPCS_2082_1_012012bib25
– start-page: 379
  year: 2016
  ident: JPCS_2082_1_012012bib5
– volume: 33
  start-page: 9259
  year: 2019
  ident: JPCS_2082_1_012012bib23
– year: 2018
  ident: JPCS_2082_1_012012bib14
– start-page: 13039
  year: 2021
  ident: JPCS_2082_1_012012bib21
– year: 2016
  ident: JPCS_2082_1_012012bib27
– volume: 33
  start-page: 9259
  year: 2019
  ident: JPCS_2082_1_012012bib16
  article-title: M2Det: A Single-Shot Object Detector Based on Multi-Level Feature Pyramid Network
  publication-title: Association for Advancement of Artificial Intelligence
– year: 2018
  ident: JPCS_2082_1_012012bib30
– start-page: 6517
  year: 2017
  ident: JPCS_2082_1_012012bib12
– start-page: 779
  year: 2016
  ident: JPCS_2082_1_012012bib11
– start-page: 2117
  year: 2017
  ident: JPCS_2082_1_012012bib20
– year: 2018
  ident: JPCS_2082_1_012012bib13
  publication-title: YOLOv3: An incremental improvement
– year: 2016
  ident: JPCS_2082_1_012012bib32
  publication-title: Deeplab: Semantic image segmentation with deep convolutional nets, atrous convolution, and fully connected crfs
– start-page: 580
  year: 2014
  ident: JPCS_2082_1_012012bib1
– ident: JPCS_2082_1_012012bib28
– start-page: 6154
  year: 2018
  ident: JPCS_2082_1_012012bib10
– start-page: 1097
  year: 2012
  ident: JPCS_2082_1_012012bib2
– start-page: 2980
  year: 2017
  ident: JPCS_2082_1_012012bib9
– volume: 34
  start-page: 1312
  year: 2012
  ident: JPCS_2082_1_012012bib7
  article-title: Cpmc: Automatic object segmentation using constrained parametric min-cuts
  publication-title: IEEE Trans. Pattern Analysis and Machine Intelligence
  doi: 10.1109/TPAMI.2011.231
– year: 2016
  ident: JPCS_2082_1_012012bib8
– start-page: 740
  year: 2014
  ident: JPCS_2082_1_012012bib34
– year: 2018
  ident: JPCS_2082_1_012012bib22
– year: 2020
  ident: JPCS_2082_1_012012bib29
– start-page: 1919
  year: 2017
  ident: JPCS_2082_1_012012bib17
– start-page: 91
  year: 2015
  ident: JPCS_2082_1_012012bib4
– volume: 111
  start-page: 98
  year: 2015
  ident: JPCS_2082_1_012012bib33
  article-title: The pascal visual object classes challenge: A retrospective
  publication-title: International Journal of Computer Vision
  doi: 10.1007/s11263-014-0733-5
– volume: 105
  start-page: 107
  year: 2019
  ident: JPCS_2082_1_012012bib26
  article-title: Gated CNN: Integrating Multi-scale Feature Layers for Object Detection
  publication-title: Pattern Recognition
– year: 2015
  ident: JPCS_2082_1_012012bib31
  publication-title: Multi-scale context aggregation by dilated convolutions
SSID ssj0033337
Score 2.283129
Snippet Feature pyramids have become an essential component in most modern object detectors, such as Mask RCNN, YOLOv3, RetinaNet. In these detectors, the pyramidal...
SourceID crossref
iop
SourceType Enrichment Source
Index Database
Publisher
StartPage 12012
Title DC-YOLOv3: A novel efficient object detection algorithm
URI https://iopscience.iop.org/article/10.1088/1742-6596/2082/1/012012
Volume 2082
hasFullText 1
inHoldings 1
isFullTextHit
isPrint
journalDatabaseRights – providerCode: PRVIOP
  databaseName: Institute of Physics Open Access Journal Titles
  customDbUrl:
  eissn: 1742-6596
  dateEnd: 99991231
  omitProxy: false
  ssIdentifier: ssj0033337
  issn: 1742-6588
  databaseCode: O3W
  dateStart: 20040101
  isFulltext: true
  titleUrlDefault: http://iopscience.iop.org/
  providerName: IOP Publishing
– providerCode: PRVPQU
  databaseName: Advanced Technologies & Aerospace Database
  customDbUrl:
  eissn: 1742-6596
  dateEnd: 99991231
  omitProxy: false
  ssIdentifier: ssj0033337
  issn: 1742-6588
  databaseCode: P5Z
  dateStart: 20040801
  isFulltext: true
  titleUrlDefault: https://search.proquest.com/hightechjournals
  providerName: ProQuest
– providerCode: PRVPQU
  databaseName: ProQuest Central
  customDbUrl:
  eissn: 1742-6596
  dateEnd: 99991231
  omitProxy: false
  ssIdentifier: ssj0033337
  issn: 1742-6588
  databaseCode: BENPR
  dateStart: 20040801
  isFulltext: true
  titleUrlDefault: https://www.proquest.com/central
  providerName: ProQuest
– providerCode: PRVPQU
  databaseName: Publicly Available Content Database
  customDbUrl:
  eissn: 1742-6596
  dateEnd: 99991231
  omitProxy: false
  ssIdentifier: ssj0033337
  issn: 1742-6588
  databaseCode: PIMPY
  dateStart: 20040801
  isFulltext: true
  titleUrlDefault: http://search.proquest.com/publiccontent
  providerName: ProQuest
link http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwrV1bS8MwFA66KfjiXZyXUdBH49YmbVPf5pyIjG14wfkUctXBbMdW9_tN2k7sgwzBPpVyTggf55bmXAA4N05GyQhhiLQnIEZEw4g0CcSedSgKCcyyQuFu2OuR4TAq1cIkk8L0X5rXvFFwDmGREEcaJob2YOBHgTm4E6_hNmz9px00XEXEeHMj0330srDGyDxhXhRpmQhZ5Hj9vlDJQ62aXfxwOLdb_7HVbbBZhJtOK-fYASsq3gXrWdqnmO2B8KYNX_vd_hxdOS0nTuZq7KisqYTxRU7C7U8aR6o0y9eKHTZ-S6aj9P1jHzzfdp7ad7AYpQAFsjYv0JJjLrTRPia5F3ETN2CluGRNwpFRQhESHTCC_SxCc825yhdEuYKxQPnaRwegEiexOgSO-aSarg6FFMJ2D2Q-Zq6wqTVRIDkKayBYwEdF0WfcjrsY0-y-mxBqUaEWFWpRoS7NUamB5jfjJG-1sZzlwkBPC7WbLSc_K5HfD9qPZQo6kfrob4segw0r13l94gmopNNPdQrWxDwdzaZ1UL3u9AYP9UwovwC_ZdR1
linkProvider IOP Publishing
linkToPdf http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwrV1ZS8NAEB7aeuCLt1jPBX005tgcG99Ka_AobUHF-rRsNhst1KS0sb_f3SQt9kGKYJ5CmFmWLzM7s8nMtwCXMsiIyMe2hmOLazYmseYTg2i2pQKKwNxmeaNw2-t0SL_v9yoQzHth0lG59F_L24IouICwLIgjusyhLc11fFdu3Imlm7rq_zQtfRTFVVhRdCXKurv4dbYiY3l5RWOkUiRkVuf1-2ALUaoqZ_Ij6ARb_zXdbdgs007UKLR2oCKSXVjLyz_5ZA-8VlN767a7U3yDGihJp2KIRE4uIWMSSkP1sQZFIsvrthLEhu_peJB9fO7DS3D73LzTyiMVNI7V2ufGUWiHPJZeyKLQ8kOZP9hChBEzSIilM3KPxC4jtpNnaqbcXzmcCJMz5gondvAB1JI0EYeA5CNhmLHHI84ViyBzbGZyVWLju1GIvTq4MwgpL_nG1bEXQ5r_9yaEKmSoQoYqZKhJC2TqYMwVRwXlxnKVKwk_Ld1vslz8YkH8odd8WpSg8u0c_W3Qc1jvtQLavu88HsOGMvWiZfEEatn4S5zCKp9mg8n4LLfNb38Q2EM
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=DC-YOLOv3%3A+A+novel+efficient+object+detection+algorithm&rft.jtitle=Journal+of+physics.+Conference+series&rft.au=Zhang%2C+Xu&rft.au=Han%2C+Fang&rft.au=Wang%2C+Ping&rft.au=Jiang%2C+Wei&rft.date=2021-11-01&rft.issn=1742-6588&rft.eissn=1742-6596&rft.volume=2082&rft.issue=1&rft.spage=12012&rft_id=info:doi/10.1088%2F1742-6596%2F2082%2F1%2F012012&rft.externalDBID=n%2Fa&rft.externalDocID=10_1088_1742_6596_2082_1_012012
thumbnail_l http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/lc.gif&issn=1742-6588&client=summon
thumbnail_m http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/mc.gif&issn=1742-6588&client=summon
thumbnail_s http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/sc.gif&issn=1742-6588&client=summon