Expanding Receptive Field YOLO for Small Object Detection
State-of-art object detection networks like YOLO, SSD and Faster R-CNN all have achieved great success in object detection. However, these algorithms have a low performance in small object detection. So, we produce the Expanding receptive field YOLO (ERF-YOLO) to deal with this problem. At first, we...
Saved in:
Published in | Journal of physics. Conference series Vol. 1314; no. 1; pp. 12202 - 12207 |
---|---|
Main Authors | , , |
Format | Journal Article |
Language | English |
Published |
Bristol
IOP Publishing
01.10.2019
|
Subjects | |
Online Access | Get full text |
Cover
Loading…
Abstract | State-of-art object detection networks like YOLO, SSD and Faster R-CNN all have achieved great success in object detection. However, these algorithms have a low performance in small object detection. So, we produce the Expanding receptive field YOLO (ERF-YOLO) to deal with this problem. At first, we propose an efficient block which is called expanding receptive field block (ERF-block) to capture more information in larger areas. Base on YOLOv2, we down-sample the low-level location information by ERF-block, and up-sample feature information by deconvolution. Then we further assemble these two parts together to make the prediction. After training the network on VOC dataset, we have a good result with 82.6% mAP (mean Average Precision) which is 4.0% higher than the original YOLOv2 network. Thanks to the efficient block, it takes 62fps to detect one image when the input size is 416×416, which could keep a real-time speed. In addition, we also evaluate the model on a remote sensing dataset which contains many small targets, and it also shows that ours model has a better performance. |
---|---|
AbstractList | Abstract
State-of-art object detection networks like YOLO, SSD and Faster R-CNN all have achieved great success in object detection. However, these algorithms have a low performance in small object detection. So, we produce the Expanding receptive field YOLO (ERF-YOLO) to deal with this problem. At first, we propose an efficient block which is called expanding receptive field block (ERF-block) to capture more information in larger areas. Base on YOLOv2, we down-sample the low-level location information by ERF-block, and up-sample feature information by deconvolution. Then we further assemble these two parts together to make the prediction. After training the network on VOC dataset, we have a good result with 82.6% mAP (mean Average Precision) which is 4.0% higher than the original YOLOv2 network. Thanks to the efficient block, it takes 62fps to detect one image when the input size is 416×416, which could keep a real-time speed. In addition, we also evaluate the model on a remote sensing dataset which contains many small targets, and it also shows that ours model has a better performance. State-of-art object detection networks like YOLO, SSD and Faster R-CNN all have achieved great success in object detection. However, these algorithms have a low performance in small object detection. So, we produce the Expanding receptive field YOLO (ERF-YOLO) to deal with this problem. At first, we propose an efficient block which is called expanding receptive field block (ERF-block) to capture more information in larger areas. Base on YOLOv2, we down-sample the low-level location information by ERF-block, and up-sample feature information by deconvolution. Then we further assemble these two parts together to make the prediction. After training the network on VOC dataset, we have a good result with 82.6% mAP (mean Average Precision) which is 4.0% higher than the original YOLOv2 network. Thanks to the efficient block, it takes 62fps to detect one image when the input size is 416×416, which could keep a real-time speed. In addition, we also evaluate the model on a remote sensing dataset which contains many small targets, and it also shows that ours model has a better performance. |
Author | Du, Zexing Yang, Jian Yin, Jinyong |
Author_xml | – sequence: 1 givenname: Zexing surname: Du fullname: Du, Zexing email: duzexing@outlook.com organization: Jiangsu Automation Research Institute , China – sequence: 2 givenname: Jinyong surname: Yin fullname: Yin, Jinyong organization: Jiangsu Automation Research Institute , China – sequence: 3 givenname: Jian surname: Yang fullname: Yang, Jian organization: Jiangsu Automation Research Institute , China |
BookMark | eNqFkF1LwzAUhoMouE1_gwHvhNp8tE1yKXPzg0HF6YVXIUkT6eiamnai_96WykQQPDfvgfO-53CeKTisfW0BOMPoEiPOY8wSEmWpyGJMcRLjGGFCEDkAk_3kcN9zfgymbbtBiPbFJkAsPhpVF2X9Ch-tsU1Xvlu4LG1VwJd8lUPnA1xvVVXBXG-s6eC17XopfX0CjpyqWnv6rTPwvFw8zW-jVX5zN79aRYYSTiIhHM00Q4oKZlxikHBcGapRUmhKLHNF_4RDiUZccGcdVSpNsFEsNVZpregMnI97m-Dfdrbt5MbvQt2flCTNmMgwErx3sdFlgm_bYJ1sQrlV4VNiJAdOciAgBxpy4CSxHDn1STomS9_8rP4_dfFH6v5hvv5tlE3h6Bdq73gd |
CitedBy_id | crossref_primary_10_1109_ACCESS_2023_3336562 crossref_primary_10_3390_electronics11172748 crossref_primary_10_1155_2022_4423744 crossref_primary_10_3390_app13127120 crossref_primary_10_3390_data9020036 crossref_primary_10_1155_2021_7918165 crossref_primary_10_1016_j_compag_2021_106219 crossref_primary_10_2174_1872212115666210827100258 crossref_primary_10_1016_j_commatsci_2021_110560 crossref_primary_10_1007_s43069_022_00163_7 |
Cites_doi | 10.1109/TPAMI.2016.2577031 |
ContentType | Journal Article |
Copyright | Published under licence by IOP Publishing Ltd 2019. This work is published under http://creativecommons.org/licenses/by/3.0/ (the “License”). Notwithstanding the ProQuest Terms and Conditions, you may use this content in accordance with the terms of the License. |
Copyright_xml | – notice: Published under licence by IOP Publishing Ltd – notice: 2019. This work is published under http://creativecommons.org/licenses/by/3.0/ (the “License”). Notwithstanding the ProQuest Terms and Conditions, you may use this content in accordance with the terms of the License. |
DBID | O3W TSCCA AAYXX CITATION 8FD 8FE 8FG ABUWG AFKRA ARAPS AZQEC BENPR BGLVJ CCPQU DWQXO H8D HCIFZ L7M P5Z P62 PIMPY PQEST PQQKQ PQUKI PRINS |
DOI | 10.1088/1742-6596/1314/1/012202 |
DatabaseName | Institute of Physics Open Access Journal Titles IOPscience (Open Access) CrossRef Technology Research Database ProQuest SciTech Collection ProQuest Technology Collection ProQuest Central (Alumni) ProQuest Central UK/Ireland Advanced Technologies & Aerospace Database (1962 - current) ProQuest Central Essentials AUTh Library subscriptions: ProQuest Central Technology Collection ProQuest One Community College ProQuest Central Aerospace Database SciTech Premium Collection (Proquest) (PQ_SDU_P3) Advanced Technologies Database with Aerospace Advanced Technologies & Aerospace Database ProQuest Advanced Technologies & Aerospace Collection Access via ProQuest (Open Access) ProQuest One Academic Eastern Edition (DO NOT USE) ProQuest One Academic ProQuest One Academic UKI Edition ProQuest Central China |
DatabaseTitle | CrossRef Publicly Available Content Database Advanced Technologies & Aerospace Collection Technology Collection Technology Research Database ProQuest Advanced Technologies & Aerospace Collection ProQuest Central Essentials ProQuest One Academic Eastern Edition ProQuest Central (Alumni Edition) SciTech Premium Collection ProQuest One Community College ProQuest Technology Collection ProQuest SciTech Collection ProQuest Central China ProQuest Central Advanced Technologies & Aerospace Database Aerospace Database ProQuest One Academic UKI Edition ProQuest Central Korea ProQuest One Academic Advanced Technologies Database with Aerospace |
DatabaseTitleList | CrossRef Publicly Available Content Database |
Database_xml | – sequence: 1 dbid: O3W name: Institute of Physics Open Access Journal Titles url: http://iopscience.iop.org/ sourceTypes: Enrichment Source Publisher – sequence: 2 dbid: 8FG name: ProQuest Technology Collection url: https://search.proquest.com/technologycollection1 sourceTypes: Aggregation Database |
DeliveryMethod | fulltext_linktorsrc |
Discipline | Physics |
DocumentTitleAlternate | Expanding Receptive Field YOLO for Small Object Detection |
EISSN | 1742-6596 |
ExternalDocumentID | 10_1088_1742_6596_1314_1_012202 JPCS_1314_1_012202 |
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 ROL SY9 T37 TR2 TSCCA UCJ W28 XSB ~02 AAYXX CITATION 8FD 8FE 8FG ABUWG AZQEC DWQXO H8D L7M P62 PQEST PQQKQ PQUKI PRINS |
ID | FETCH-LOGICAL-c3282-99f36b70a397cf4c09f8ac3b04db32e7fd088f04b0898fef3aa541ca75ceabba3 |
IEDL.DBID | O3W |
ISSN | 1742-6588 |
IngestDate | Thu Oct 10 17:30:47 EDT 2024 Fri Aug 23 00:44:40 EDT 2024 Thu Jan 07 13:53:39 EST 2021 Wed Aug 21 03:33:11 EDT 2024 |
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-c3282-99f36b70a397cf4c09f8ac3b04db32e7fd088f04b0898fef3aa541ca75ceabba3 |
OpenAccessLink | https://iopscience.iop.org/article/10.1088/1742-6596/1314/1/012202 |
PQID | 2567961098 |
PQPubID | 4998668 |
PageCount | 6 |
ParticipantIDs | iop_journals_10_1088_1742_6596_1314_1_012202 crossref_primary_10_1088_1742_6596_1314_1_012202 proquest_journals_2567961098 |
PublicationCentury | 2000 |
PublicationDate | 20191001 |
PublicationDateYYYYMMDD | 2019-10-01 |
PublicationDate_xml | – month: 10 year: 2019 text: 20191001 day: 01 |
PublicationDecade | 2010 |
PublicationPlace | Bristol |
PublicationPlace_xml | – name: Bristol |
PublicationTitle | Journal of physics. Conference series |
PublicationTitleAlternate | J. Phys.: Conf. Ser |
PublicationYear | 2019 |
Publisher | IOP Publishing |
Publisher_xml | – name: IOP Publishing |
References | Liu (JPCS_1314_1_012202bib4) 2016 Lin (JPCS_1314_1_012202bib8) 2017 JPCS_1314_1_012202bib6 JPCS_1314_1_012202bib5 JPCS_1314_1_012202bib7 Redmon (JPCS_1314_1_012202bib3) 2017 Ren (JPCS_1314_1_012202bib1) 2017; 39 JPCS_1314_1_012202bib14 Maoke (JPCS_1314_1_012202bib9) 2018 JPCS_1314_1_012202bib15 JPCS_1314_1_012202bib12 JPCS_1314_1_012202bib13 Redmon (JPCS_1314_1_012202bib2) 2016 JPCS_1314_1_012202bib10 JPCS_1314_1_012202bib11 |
References_xml | – start-page: 779 year: 2016 ident: JPCS_1314_1_012202bib2 article-title: You Only Look Once: Unified, Real-Time Object Detection contributor: fullname: Redmon – start-page: 6517 year: 2017 ident: JPCS_1314_1_012202bib3 article-title: YOLO9000: Better, Faster, Stronger contributor: fullname: Redmon – start-page: 21 year: 2016 ident: JPCS_1314_1_012202bib4 article-title: SSD: Single Shot MultiBox Detector contributor: fullname: Liu – start-page: 2117 year: 2017 ident: JPCS_1314_1_012202bib8 article-title: Feature pyramid networks for object detection (2017) contributor: fullname: Lin – ident: JPCS_1314_1_012202bib5 – ident: JPCS_1314_1_012202bib6 – volume: 39 start-page: 1137 year: 2017 ident: JPCS_1314_1_012202bib1 article-title: Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks publication-title: IEEE Transactions on Pattern Analysis and Machine Intelligence doi: 10.1109/TPAMI.2016.2577031 contributor: fullname: Ren – ident: JPCS_1314_1_012202bib10 – ident: JPCS_1314_1_012202bib11 – ident: JPCS_1314_1_012202bib7 – ident: JPCS_1314_1_012202bib13 – ident: JPCS_1314_1_012202bib12 – ident: JPCS_1314_1_012202bib15 – ident: JPCS_1314_1_012202bib14 – start-page: 3684 year: 2018 ident: JPCS_1314_1_012202bib9 article-title: DenseASPP for Semantic Segmentation in Street Scenes contributor: fullname: Maoke |
SSID | ssj0033337 |
Score | 2.3855686 |
Snippet | State-of-art object detection networks like YOLO, SSD and Faster R-CNN all have achieved great success in object detection. However, these algorithms have a... Abstract State-of-art object detection networks like YOLO, SSD and Faster R-CNN all have achieved great success in object detection. However, these algorithms... |
SourceID | proquest crossref iop |
SourceType | Aggregation Database Enrichment Source Publisher |
StartPage | 12202 |
SubjectTerms | Algorithms Datasets Object recognition Physics Remote sensing |
SummonAdditionalLinks | – databaseName: AUTh Library subscriptions: ProQuest Central dbid: BENPR link: http://utb.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwhV3dS8MwEA86EXwRP3E6JaCPlrVN2iRPonNjiLrhB-hTyCcIs6tuin--SZsyRNA-tveSX5O7y93v7gA4SZmMM2vyCOUER9g51JEg7lzRHEmdUZqTqlD45jYfPuKrp-wpBNxmgVbZ6MRKUeup8jHyrjPNhPne4PSsfIv81CifXQ0jNJbBSppgn6Zduejfju8aXYzcQ-qSyDRytpY2DC937QvvWN5NUIK7SdcnmUJ0pbFPyy_T8peSrizPYAOsB5cRntf_eBMsmWILrFbUTTXbBqz_Vda1KdD5gJ6k8mngwBPT4PPoegSdVwrvX8VkAkfSB13gpZlX_KtiBzwO-g-9YRQGIkQKuatRxJhFuSSxcE6EsljFzFKhkIyxlig1xGq3MBtjGVNGrbFIiAwnSpBMGSGlQLugVUwLswdgFmuDnelmWhGsfAcbjYhGmKS-o1eG2iBuYOBl3feCV_lqSrlHjnvkuEeOJ7xGrg1OHVw8nIHZ_-LHP8Svxr37nxK81LYNOg36C9HFXtj_-_MBWHPrYTX5rgNa8_cPc-iciLk8CjvlG2ayu7s priority: 102 providerName: ProQuest |
Title | Expanding Receptive Field YOLO for Small Object Detection |
URI | https://iopscience.iop.org/article/10.1088/1742-6596/1314/1/012202 https://www.proquest.com/docview/2567961098 |
Volume | 1314 |
hasFullText | 1 |
inHoldings | 1 |
isFullTextHit | |
isPrint | |
link | http://utb.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwlV1dS8MwFA26IfgifuJ0joA-WtsuaZM86lz9QO3wA_UpJGnyNOfQKf58b9oOHSJin1q4bZNDk3vanHuK0F5X6ChxNg1IymhAgVAHisG44inRRcJ5yspC4cur9PSOnj8kD99rYZ7H9dR_ALuVUXAFYS2I4yFw6G6QJiINYxLTMA796pD3k2xC8i1JUk7up7MxgY1VRZH-JM6nGq_fLzSToeahFT-m6TL3ZMtoqSaN-LBq4gqas6NVtFCKN83rGhL9j3FVnYKBBXqZyrvFmZem4cf8IsfAS_HNkxoOca79Zxd8bCelAmu0ju6y_m3vNKh_iRAYAi9HgRCOpJpFCmiEcdREwnFliI5ooUnXMldAx1xEdcQFd9YRpRIaG8USY5XWimygxuh5ZDcRTqLCUkjeojCMGu9hUxBWEMq63tMrIS0UTWGQ48r5QpYr1pxLj5z0yEmPnIxlhVwL7QNcsh4Fr3-H786Enw96N7MRcly4FmpP0f8KBaLGhHeK51v_u-c2WoT-iUqO10aNycub3QFaMdEdNM-zkw5qHvWvBtdwdJYPOuWz9Akiqb30 |
link.rule.ids | 315,786,790,12792,21416,27955,27956,33406,33777,38898,38923,43633,43838,53875,53901 |
linkProvider | IOP Publishing |
linkToHtml | http://utb.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwhV1LSwMxEA62InoRn1itGtCjS3eb7CY5idTWWqsVqlBPIU8QarvaKv58k31QiqB73J1Lvp3MTCbfzABw3mQyjK1JApQQHGAXUAeCuH1FEyR1TGlCskLh-4ek-4x7o3hUJNxmBa2ytImZodZT5XPkDeeaCfO9well-h74qVH-drUYoVEBqxglyOs57dyUlhi5h-QFkc3AeVpa8rvcoa94x5JGhCLciBr-iqnIrZTeqfI6TX-Z6MzvdLbAZhEwwqv8D2-DFTPZAWsZcVPNdgFrf6d5ZQp0EaCnqHwZ2PG0NPgy6A-gi0nh8E2Mx3AgfcoFXpt5xr6a7IHnTvup1Q2KcQiBQu5gFDBmUSJJKFwIoSxWIbNUKCRDrCVqGmK1W5gNsQwpo9ZYJESMIyVIrIyQUqB9UJ1MJ-YAwDjUBjvHzbQiWPn-NRoRjTBp-n5eMaqBsISBp3nXC57dVlPKPXLcI8c9cjziOXI1cOHg4sUOmP0vfrYk3ntsDZcleKptDdRL9BeiC004_PvzKVjvPt33ef_24e4IbLi1sZyGVwfV-cenOXbhxFyeZDrzAwuDvUI |
linkToPdf | http://utb.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwpV1RT9UwFD4BjMYXRdR4FbEJPLq7bu3W9tEACyByb4Ik-NS0XeuDeLmRizH-ek_XTXMxhBj3tIfTrf22c_p1_c4ZwE6pLK2CrzNWC55xJNSZEehXsma2raSsRZco_OGkPjjjR-fV-Qo0v3NhLud96B_jaSoUnCDsBXEyRw5dZnWl6rxgBc-LPO4O0TKft2EV7qEDiyjtO5xMh4jM8BApMTI2lHLQed1-saVZahV78leo7uaf5jF8HnqeZCdfxtcLO3Y_bxR1_P-hrcOjnqKSd6nVE1jxsw2430lF3dVTUPs_5ikXhiDnjKKY7540UQhHPk2OJwRZMDn9ai4uyMTGjzxkzy86vdfsGZw1-x93D7L-BwyZY7gUy5QKrLaCGiQtLnBHVZDGMUt5a1npRWix04FyS6WSwQdmTMULZ0TlvLHWsOewNruc-RdAKtp6jlRBtU5wFyvmtEy0jIsyVhCr2AjoALiepzobutsfl1JHVHRERUdUdKETKiN4izjq3ueu7jbfXjI_mu6eLltohHkEm8Nz_mOKtFCoWJdevvy3e76BB9O9Rh8fnrx_BQ9xqCrpADdhbfHt2r9GPrOwW93L-gsqpt_V |
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=Expanding+Receptive+Field+YOLO+for+Small+Object+Detection&rft.jtitle=Journal+of+physics.+Conference+series&rft.au=Du%2C+Zexing&rft.au=Yin%2C+Jinyong&rft.au=Yang%2C+Jian&rft.date=2019-10-01&rft.issn=1742-6588&rft.eissn=1742-6596&rft.volume=1314&rft.issue=1&rft.spage=12202&rft_id=info:doi/10.1088%2F1742-6596%2F1314%2F1%2F012202&rft.externalDBID=n%2Fa&rft.externalDocID=10_1088_1742_6596_1314_1_012202 |
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 |