Loading…

Deep Multiphase Level Set for Scene Parsing

Recently, Fully Convolutional Network (FCN) seems to be the go-to architecture for image segmentation, including semantic scene parsing. However, it is difficult for a generic FCN to predict semantic labels around the object boundaries, thus FCN-based methods usually produce parsing results with ina...

Full description

Saved in:
Bibliographic Details
Published in:IEEE transactions on image processing 2020-01, Vol.29, p.4556-4567
Main Authors: Zhang, Pingping, Liu, Wei, Lei, Yinjie, Wang, Hongyu, Lu, Huchuan
Format: Article
Language:English
Subjects:
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
cited_by cdi_FETCH-LOGICAL-c347t-8d8e40f000c39a4d8932e453f53f5ab21aea063eaf59da5912494ce9bfb0fbec3
cites cdi_FETCH-LOGICAL-c347t-8d8e40f000c39a4d8932e453f53f5ab21aea063eaf59da5912494ce9bfb0fbec3
container_end_page 4567
container_issue
container_start_page 4556
container_title IEEE transactions on image processing
container_volume 29
creator Zhang, Pingping
Liu, Wei
Lei, Yinjie
Wang, Hongyu
Lu, Huchuan
description Recently, Fully Convolutional Network (FCN) seems to be the go-to architecture for image segmentation, including semantic scene parsing. However, it is difficult for a generic FCN to predict semantic labels around the object boundaries, thus FCN-based methods usually produce parsing results with inaccurate boundaries. Meanwhile, many works have demonstrate that level set based active contours are superior to the boundary estimation in sub-pixel accuracy. However, they are quite sensitive to initial settings. To address these limitations, in this paper we propose a novel Deep Multiphase Level Set (DMLS) method for semantic scene parsing, which efficiently incorporates multiphase level sets into deep neural networks. The proposed method consists of three modules, i.e., recurrent FCNs, adaptive multiphase level set, and deeply supervised learning. More specifically, recurrent FCNs learn multi-level representations of input images with different contexts. Adaptive multiphase level set drives the discriminative contour for each semantic class, which makes use of the advantages of both global and local information. In each time-step of the recurrent FCNs, deeply supervised learning is incorporated for model training. Extensive experiments on three public benchmarks have shown that our proposed method achieves new state-of-the-art performances. The source codes will be released at https://github.com/Pchank/DMLS-for-SSP.
doi_str_mv 10.1109/TIP.2019.2957915
format article
fullrecord <record><control><sourceid>proquest_cross</sourceid><recordid>TN_cdi_crossref_primary_10_1109_TIP_2019_2957915</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><ieee_id>9003517</ieee_id><sourcerecordid>2362081921</sourcerecordid><originalsourceid>FETCH-LOGICAL-c347t-8d8e40f000c39a4d8932e453f53f5ab21aea063eaf59da5912494ce9bfb0fbec3</originalsourceid><addsrcrecordid>eNpdkEtLw0AQgBdRbK3eBUECXgRJnX012aPUV6FiofUcNsmspqRJ3E0E_70bW3sQZpiB-WYYPkLOKYwpBXW7mi3GDKgaMyUjReUBGVIlaAgg2KHvQUZhRIUakBPn1gBUSDo5JgPOIJ74HJKbe8QmeOnKtmg-tMNgjl9YBktsA1PbYJlhhcFCW1dU76fkyOjS4dmujsjb48Nq-hzOX59m07t5mHERtWGcxyjAAEDGlRZ5rDhDIbnpQ6eMatQw4aiNVLmWijKhRIYqNSmYFDM-Itfbu42tPzt0bbIpXIZlqSusO5cw3v9OFaMevfqHruvOVv67X0qAikTkKdhSma2ds2iSxhYbbb8TCkkvMvEik15kshPpVy53h7t0g_l-4c-cBy62QIGI-7EC4JJG_AfWkXRY</addsrcrecordid><sourcetype>Aggregation Database</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype><pqid>2362409747</pqid></control><display><type>article</type><title>Deep Multiphase Level Set for Scene Parsing</title><source>IEEE Electronic Library (IEL) Journals</source><creator>Zhang, Pingping ; Liu, Wei ; Lei, Yinjie ; Wang, Hongyu ; Lu, Huchuan</creator><creatorcontrib>Zhang, Pingping ; Liu, Wei ; Lei, Yinjie ; Wang, Hongyu ; Lu, Huchuan</creatorcontrib><description>Recently, Fully Convolutional Network (FCN) seems to be the go-to architecture for image segmentation, including semantic scene parsing. However, it is difficult for a generic FCN to predict semantic labels around the object boundaries, thus FCN-based methods usually produce parsing results with inaccurate boundaries. Meanwhile, many works have demonstrate that level set based active contours are superior to the boundary estimation in sub-pixel accuracy. However, they are quite sensitive to initial settings. To address these limitations, in this paper we propose a novel Deep Multiphase Level Set (DMLS) method for semantic scene parsing, which efficiently incorporates multiphase level sets into deep neural networks. The proposed method consists of three modules, i.e., recurrent FCNs, adaptive multiphase level set, and deeply supervised learning. More specifically, recurrent FCNs learn multi-level representations of input images with different contexts. Adaptive multiphase level set drives the discriminative contour for each semantic class, which makes use of the advantages of both global and local information. In each time-step of the recurrent FCNs, deeply supervised learning is incorporated for model training. Extensive experiments on three public benchmarks have shown that our proposed method achieves new state-of-the-art performances. The source codes will be released at https://github.com/Pchank/DMLS-for-SSP.</description><identifier>ISSN: 1057-7149</identifier><identifier>EISSN: 1941-0042</identifier><identifier>DOI: 10.1109/TIP.2019.2957915</identifier><identifier>PMID: 32086208</identifier><identifier>CODEN: IIPRE4</identifier><language>eng</language><publisher>United States: IEEE</publisher><subject>Active contours ; Artificial neural networks ; Boundaries ; DSL ; Feature extraction ; Image segmentation ; Level set ; Machine learning ; Multiphase ; multiphase level set ; object boundary estimation ; Production methods ; recurrent convolutional network ; Semantic scene parsing ; Semantics ; Supervised learning</subject><ispartof>IEEE transactions on image processing, 2020-01, Vol.29, p.4556-4567</ispartof><rights>Copyright The Institute of Electrical and Electronics Engineers, Inc. (IEEE) 2020</rights><lds50>peer_reviewed</lds50><woscitedreferencessubscribed>false</woscitedreferencessubscribed><citedby>FETCH-LOGICAL-c347t-8d8e40f000c39a4d8932e453f53f5ab21aea063eaf59da5912494ce9bfb0fbec3</citedby><cites>FETCH-LOGICAL-c347t-8d8e40f000c39a4d8932e453f53f5ab21aea063eaf59da5912494ce9bfb0fbec3</cites><orcidid>0000-0002-1038-412X ; 0000-0002-6668-9758 ; 0000-0001-6351-9019 ; 0000-0001-6856-3342</orcidid></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><linktohtml>$$Uhttps://ieeexplore.ieee.org/document/9003517$$EHTML$$P50$$Gieee$$H</linktohtml><link.rule.ids>314,776,780,27903,27904,54774</link.rule.ids><backlink>$$Uhttps://www.ncbi.nlm.nih.gov/pubmed/32086208$$D View this record in MEDLINE/PubMed$$Hfree_for_read</backlink></links><search><creatorcontrib>Zhang, Pingping</creatorcontrib><creatorcontrib>Liu, Wei</creatorcontrib><creatorcontrib>Lei, Yinjie</creatorcontrib><creatorcontrib>Wang, Hongyu</creatorcontrib><creatorcontrib>Lu, Huchuan</creatorcontrib><title>Deep Multiphase Level Set for Scene Parsing</title><title>IEEE transactions on image processing</title><addtitle>TIP</addtitle><addtitle>IEEE Trans Image Process</addtitle><description>Recently, Fully Convolutional Network (FCN) seems to be the go-to architecture for image segmentation, including semantic scene parsing. However, it is difficult for a generic FCN to predict semantic labels around the object boundaries, thus FCN-based methods usually produce parsing results with inaccurate boundaries. Meanwhile, many works have demonstrate that level set based active contours are superior to the boundary estimation in sub-pixel accuracy. However, they are quite sensitive to initial settings. To address these limitations, in this paper we propose a novel Deep Multiphase Level Set (DMLS) method for semantic scene parsing, which efficiently incorporates multiphase level sets into deep neural networks. The proposed method consists of three modules, i.e., recurrent FCNs, adaptive multiphase level set, and deeply supervised learning. More specifically, recurrent FCNs learn multi-level representations of input images with different contexts. Adaptive multiphase level set drives the discriminative contour for each semantic class, which makes use of the advantages of both global and local information. In each time-step of the recurrent FCNs, deeply supervised learning is incorporated for model training. Extensive experiments on three public benchmarks have shown that our proposed method achieves new state-of-the-art performances. The source codes will be released at https://github.com/Pchank/DMLS-for-SSP.</description><subject>Active contours</subject><subject>Artificial neural networks</subject><subject>Boundaries</subject><subject>DSL</subject><subject>Feature extraction</subject><subject>Image segmentation</subject><subject>Level set</subject><subject>Machine learning</subject><subject>Multiphase</subject><subject>multiphase level set</subject><subject>object boundary estimation</subject><subject>Production methods</subject><subject>recurrent convolutional network</subject><subject>Semantic scene parsing</subject><subject>Semantics</subject><subject>Supervised learning</subject><issn>1057-7149</issn><issn>1941-0042</issn><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2020</creationdate><recordtype>article</recordtype><recordid>eNpdkEtLw0AQgBdRbK3eBUECXgRJnX012aPUV6FiofUcNsmspqRJ3E0E_70bW3sQZpiB-WYYPkLOKYwpBXW7mi3GDKgaMyUjReUBGVIlaAgg2KHvQUZhRIUakBPn1gBUSDo5JgPOIJ74HJKbe8QmeOnKtmg-tMNgjl9YBktsA1PbYJlhhcFCW1dU76fkyOjS4dmujsjb48Nq-hzOX59m07t5mHERtWGcxyjAAEDGlRZ5rDhDIbnpQ6eMatQw4aiNVLmWijKhRIYqNSmYFDM-Itfbu42tPzt0bbIpXIZlqSusO5cw3v9OFaMevfqHruvOVv67X0qAikTkKdhSma2ds2iSxhYbbb8TCkkvMvEik15kshPpVy53h7t0g_l-4c-cBy62QIGI-7EC4JJG_AfWkXRY</recordid><startdate>20200101</startdate><enddate>20200101</enddate><creator>Zhang, Pingping</creator><creator>Liu, Wei</creator><creator>Lei, Yinjie</creator><creator>Wang, Hongyu</creator><creator>Lu, Huchuan</creator><general>IEEE</general><general>The Institute of Electrical and Electronics Engineers, Inc. (IEEE)</general><scope>97E</scope><scope>RIA</scope><scope>RIE</scope><scope>NPM</scope><scope>AAYXX</scope><scope>CITATION</scope><scope>7SC</scope><scope>7SP</scope><scope>8FD</scope><scope>JQ2</scope><scope>L7M</scope><scope>L~C</scope><scope>L~D</scope><scope>7X8</scope><orcidid>https://orcid.org/0000-0002-1038-412X</orcidid><orcidid>https://orcid.org/0000-0002-6668-9758</orcidid><orcidid>https://orcid.org/0000-0001-6351-9019</orcidid><orcidid>https://orcid.org/0000-0001-6856-3342</orcidid></search><sort><creationdate>20200101</creationdate><title>Deep Multiphase Level Set for Scene Parsing</title><author>Zhang, Pingping ; Liu, Wei ; Lei, Yinjie ; Wang, Hongyu ; Lu, Huchuan</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-c347t-8d8e40f000c39a4d8932e453f53f5ab21aea063eaf59da5912494ce9bfb0fbec3</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2020</creationdate><topic>Active contours</topic><topic>Artificial neural networks</topic><topic>Boundaries</topic><topic>DSL</topic><topic>Feature extraction</topic><topic>Image segmentation</topic><topic>Level set</topic><topic>Machine learning</topic><topic>Multiphase</topic><topic>multiphase level set</topic><topic>object boundary estimation</topic><topic>Production methods</topic><topic>recurrent convolutional network</topic><topic>Semantic scene parsing</topic><topic>Semantics</topic><topic>Supervised learning</topic><toplevel>peer_reviewed</toplevel><toplevel>online_resources</toplevel><creatorcontrib>Zhang, Pingping</creatorcontrib><creatorcontrib>Liu, Wei</creatorcontrib><creatorcontrib>Lei, Yinjie</creatorcontrib><creatorcontrib>Wang, Hongyu</creatorcontrib><creatorcontrib>Lu, Huchuan</creatorcontrib><collection>IEEE All-Society Periodicals Package (ASPP) 2005-present</collection><collection>IEEE All-Society Periodicals Package (ASPP) 1998-Present</collection><collection>IEEE Electronic Library (IEL)</collection><collection>PubMed</collection><collection>CrossRef</collection><collection>Computer and Information Systems Abstracts</collection><collection>Electronics &amp; Communications Abstracts</collection><collection>Technology Research Database</collection><collection>ProQuest Computer Science Collection</collection><collection>Advanced Technologies Database with Aerospace</collection><collection>Computer and Information Systems Abstracts – Academic</collection><collection>Computer and Information Systems Abstracts Professional</collection><collection>MEDLINE - Academic</collection><jtitle>IEEE transactions on image processing</jtitle></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext</fulltext></delivery><addata><au>Zhang, Pingping</au><au>Liu, Wei</au><au>Lei, Yinjie</au><au>Wang, Hongyu</au><au>Lu, Huchuan</au><format>journal</format><genre>article</genre><ristype>JOUR</ristype><atitle>Deep Multiphase Level Set for Scene Parsing</atitle><jtitle>IEEE transactions on image processing</jtitle><stitle>TIP</stitle><addtitle>IEEE Trans Image Process</addtitle><date>2020-01-01</date><risdate>2020</risdate><volume>29</volume><spage>4556</spage><epage>4567</epage><pages>4556-4567</pages><issn>1057-7149</issn><eissn>1941-0042</eissn><coden>IIPRE4</coden><abstract>Recently, Fully Convolutional Network (FCN) seems to be the go-to architecture for image segmentation, including semantic scene parsing. However, it is difficult for a generic FCN to predict semantic labels around the object boundaries, thus FCN-based methods usually produce parsing results with inaccurate boundaries. Meanwhile, many works have demonstrate that level set based active contours are superior to the boundary estimation in sub-pixel accuracy. However, they are quite sensitive to initial settings. To address these limitations, in this paper we propose a novel Deep Multiphase Level Set (DMLS) method for semantic scene parsing, which efficiently incorporates multiphase level sets into deep neural networks. The proposed method consists of three modules, i.e., recurrent FCNs, adaptive multiphase level set, and deeply supervised learning. More specifically, recurrent FCNs learn multi-level representations of input images with different contexts. Adaptive multiphase level set drives the discriminative contour for each semantic class, which makes use of the advantages of both global and local information. In each time-step of the recurrent FCNs, deeply supervised learning is incorporated for model training. Extensive experiments on three public benchmarks have shown that our proposed method achieves new state-of-the-art performances. The source codes will be released at https://github.com/Pchank/DMLS-for-SSP.</abstract><cop>United States</cop><pub>IEEE</pub><pmid>32086208</pmid><doi>10.1109/TIP.2019.2957915</doi><tpages>12</tpages><orcidid>https://orcid.org/0000-0002-1038-412X</orcidid><orcidid>https://orcid.org/0000-0002-6668-9758</orcidid><orcidid>https://orcid.org/0000-0001-6351-9019</orcidid><orcidid>https://orcid.org/0000-0001-6856-3342</orcidid></addata></record>
fulltext fulltext
identifier ISSN: 1057-7149
ispartof IEEE transactions on image processing, 2020-01, Vol.29, p.4556-4567
issn 1057-7149
1941-0042
language eng
recordid cdi_crossref_primary_10_1109_TIP_2019_2957915
source IEEE Electronic Library (IEL) Journals
subjects Active contours
Artificial neural networks
Boundaries
DSL
Feature extraction
Image segmentation
Level set
Machine learning
Multiphase
multiphase level set
object boundary estimation
Production methods
recurrent convolutional network
Semantic scene parsing
Semantics
Supervised learning
title Deep Multiphase Level Set for Scene Parsing
url http://sfxeu10.hosted.exlibrisgroup.com/loughborough?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2025-01-24T07%3A53%3A06IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-proquest_cross&rft_val_fmt=info:ofi/fmt:kev:mtx:journal&rft.genre=article&rft.atitle=Deep%20Multiphase%20Level%20Set%20for%20Scene%20Parsing&rft.jtitle=IEEE%20transactions%20on%20image%20processing&rft.au=Zhang,%20Pingping&rft.date=2020-01-01&rft.volume=29&rft.spage=4556&rft.epage=4567&rft.pages=4556-4567&rft.issn=1057-7149&rft.eissn=1941-0042&rft.coden=IIPRE4&rft_id=info:doi/10.1109/TIP.2019.2957915&rft_dat=%3Cproquest_cross%3E2362081921%3C/proquest_cross%3E%3Cgrp_id%3Ecdi_FETCH-LOGICAL-c347t-8d8e40f000c39a4d8932e453f53f5ab21aea063eaf59da5912494ce9bfb0fbec3%3C/grp_id%3E%3Coa%3E%3C/oa%3E%3Curl%3E%3C/url%3E&rft_id=info:oai/&rft_pqid=2362409747&rft_id=info:pmid/32086208&rft_ieee_id=9003517&rfr_iscdi=true