{"id":"https://openalex.org/W3114304902","doi":"https://doi.org/10.1109/tmi.2020.3046692","title":"Analyzing Overfitting Under Class Imbalance in Neural Networks for Image Segmentation","display_name":"Analyzing Overfitting Under Class Imbalance in Neural Networks for Image Segmentation","publication_year":2020,"publication_date":"2020-12-22","ids":{"openalex":"https://openalex.org/W3114304902","doi":"https://doi.org/10.1109/tmi.2020.3046692","mag":"3114304902","pmid":"https://pubmed.ncbi.nlm.nih.gov/33351758"},"language":"en","primary_location":{"id":"doi:10.1109/tmi.2020.3046692","is_oa":false,"landing_page_url":"https://doi.org/10.1109/tmi.2020.3046692","pdf_url":null,"source":{"id":"https://openalex.org/S58069681","display_name":"IEEE Transactions on Medical Imaging","issn_l":"0278-0062","issn":["0278-0062","1558-254X"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310319808","host_organization_name":"Institute of Electrical and Electronics Engineers","host_organization_lineage":["https://openalex.org/P4310319808"],"host_organization_lineage_names":["Institute of Electrical and Electronics Engineers"],"type":"journal"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"IEEE Transactions on Medical Imaging","raw_type":"journal-article"},"type":"article","indexed_in":["arxiv","crossref","pubmed"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://arxiv.org/pdf/2102.10365","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":null,"display_name":"Zeju Li","orcid":"https://orcid.org/0000-0002-4608-2959"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zeju Li","raw_affiliation_strings":[],"raw_orcid":"https://orcid.org/0000-0002-4608-2959","affiliations":[]},{"author_position":"middle","author":{"id":null,"display_name":"Konstantinos Kamnitsas","orcid":"https://orcid.org/0000-0003-3281-6509"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Konstantinos Kamnitsas","raw_affiliation_strings":[],"raw_orcid":"https://orcid.org/0000-0003-3281-6509","affiliations":[]},{"author_position":"last","author":{"id":null,"display_name":"Ben Glocker","orcid":"https://orcid.org/0000-0002-4897-9356"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Ben Glocker","raw_affiliation_strings":[],"raw_orcid":"https://orcid.org/0000-0002-4897-9356","affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":0,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":9.1263,"has_fulltext":false,"cited_by_count":188,"citation_normalized_percentile":{"value":0.98357895,"is_in_top_1_percent":false,"is_in_top_10_percent":true},"cited_by_percentile_year":{"min":97,"max":100},"biblio":{"volume":"40","issue":"3","first_page":"1065","last_page":"1077"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T11652","display_name":"Imbalanced Data Classification Techniques","score":0.5511000156402588,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},"topics":[{"id":"https://openalex.org/T11652","display_name":"Imbalanced Data Classification Techniques","score":0.5511000156402588,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T11689","display_name":"Adversarial Robustness in Machine Learning","score":0.1598999947309494,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T11307","display_name":"Domain Adaptation and Few-Shot Learning","score":0.03759999945759773,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/overfitting","display_name":"Overfitting","score":0.9229000210762024},{"id":"https://openalex.org/keywords/segmentation","display_name":"Segmentation","score":0.6717000007629395},{"id":"https://openalex.org/keywords/margin","display_name":"Margin (machine learning)","score":0.6640999913215637},{"id":"https://openalex.org/keywords/pattern-recognition","display_name":"Pattern recognition (psychology)","score":0.5837000012397766},{"id":"https://openalex.org/keywords/image-segmentation","display_name":"Image segmentation","score":0.5497000217437744},{"id":"https://openalex.org/keywords/regularization","display_name":"Regularization (linguistics)","score":0.5097000002861023},{"id":"https://openalex.org/keywords/logit","display_name":"Logit","score":0.5044000148773193},{"id":"https://openalex.org/keywords/artificial-neural-network","display_name":"Artificial neural network","score":0.47760000824928284}],"concepts":[{"id":"https://openalex.org/C22019652","wikidata":"https://www.wikidata.org/wiki/Q331309","display_name":"Overfitting","level":3,"score":0.9229000210762024},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.7554000020027161},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7037000060081482},{"id":"https://openalex.org/C89600930","wikidata":"https://www.wikidata.org/wiki/Q1423946","display_name":"Segmentation","level":2,"score":0.6717000007629395},{"id":"https://openalex.org/C774472","wikidata":"https://www.wikidata.org/wiki/Q6760393","display_name":"Margin (machine learning)","level":2,"score":0.6640999913215637},{"id":"https://openalex.org/C153180895","wikidata":"https://www.wikidata.org/wiki/Q7148389","display_name":"Pattern recognition (psychology)","level":2,"score":0.5837000012397766},{"id":"https://openalex.org/C124504099","wikidata":"https://www.wikidata.org/wiki/Q56933","display_name":"Image segmentation","level":3,"score":0.5497000217437744},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.5184999704360962},{"id":"https://openalex.org/C2776135515","wikidata":"https://www.wikidata.org/wiki/Q17143721","display_name":"Regularization (linguistics)","level":2,"score":0.5097000002861023},{"id":"https://openalex.org/C140331021","wikidata":"https://www.wikidata.org/wiki/Q1868104","display_name":"Logit","level":2,"score":0.5044000148773193},{"id":"https://openalex.org/C50644808","wikidata":"https://www.wikidata.org/wiki/Q192776","display_name":"Artificial neural network","level":2,"score":0.47760000824928284},{"id":"https://openalex.org/C2777212361","wikidata":"https://www.wikidata.org/wiki/Q5127848","display_name":"Class (philosophy)","level":2,"score":0.4758000075817108},{"id":"https://openalex.org/C115961682","wikidata":"https://www.wikidata.org/wiki/Q860623","display_name":"Image (mathematics)","level":2,"score":0.3718000054359436},{"id":"https://openalex.org/C151956035","wikidata":"https://www.wikidata.org/wiki/Q1132755","display_name":"Logistic regression","level":2,"score":0.3671000003814697},{"id":"https://openalex.org/C2984842247","wikidata":"https://www.wikidata.org/wiki/Q197536","display_name":"Deep neural networks","level":3,"score":0.31439998745918274},{"id":"https://openalex.org/C75294576","wikidata":"https://www.wikidata.org/wiki/Q5165192","display_name":"Contextual image classification","level":3,"score":0.31200000643730164},{"id":"https://openalex.org/C108583219","wikidata":"https://www.wikidata.org/wiki/Q197536","display_name":"Deep learning","level":2,"score":0.2955000102519989},{"id":"https://openalex.org/C65885262","wikidata":"https://www.wikidata.org/wiki/Q7429708","display_name":"Scale-space segmentation","level":4,"score":0.2922999858856201},{"id":"https://openalex.org/C37736160","wikidata":"https://www.wikidata.org/wiki/Q1801315","display_name":"Adversarial system","level":2,"score":0.2687000036239624},{"id":"https://openalex.org/C125308379","wikidata":"https://www.wikidata.org/wiki/Q363057","display_name":"Market segmentation","level":2,"score":0.26739999651908875},{"id":"https://openalex.org/C52622490","wikidata":"https://www.wikidata.org/wiki/Q1026626","display_name":"Feature extraction","level":2,"score":0.259799987077713},{"id":"https://openalex.org/C2776502983","wikidata":"https://www.wikidata.org/wiki/Q690182","display_name":"Contrast (vision)","level":2,"score":0.25450000166893005}],"mesh":[{"descriptor_ui":"D007091","descriptor_name":"Image Processing, Computer-Assisted","qualifier_ui":null,"qualifier_name":null,"is_major_topic":true},{"descriptor_ui":"D007091","descriptor_name":"Image Processing, Computer-Assisted","qualifier_ui":null,"qualifier_name":null,"is_major_topic":true},{"descriptor_ui":"D007091","descriptor_name":"Image Processing, Computer-Assisted","qualifier_ui":null,"qualifier_name":null,"is_major_topic":true},{"descriptor_ui":"D016208","descriptor_name":"Databases, Factual","qualifier_ui":null,"qualifier_name":null,"is_major_topic":false},{"descriptor_ui":"D016208","descriptor_name":"Databases, Factual","qualifier_ui":null,"qualifier_name":null,"is_major_topic":false},{"descriptor_ui":"D016208","descriptor_name":"Databases, Factual","qualifier_ui":null,"qualifier_name":null,"is_major_topic":false},{"descriptor_ui":"D016571","descriptor_name":"Neural Networks, Computer","qualifier_ui":null,"qualifier_name":null,"is_major_topic":true},{"descriptor_ui":"D016571","descriptor_name":"Neural Networks, Computer","qualifier_ui":null,"qualifier_name":null,"is_major_topic":true},{"descriptor_ui":"D016571","descriptor_name":"Neural Networks, Computer","qualifier_ui":null,"qualifier_name":null,"is_major_topic":true}],"locations_count":5,"locations":[{"id":"doi:10.1109/tmi.2020.3046692","is_oa":false,"landing_page_url":"https://doi.org/10.1109/tmi.2020.3046692","pdf_url":null,"source":{"id":"https://openalex.org/S58069681","display_name":"IEEE Transactions on Medical Imaging","issn_l":"0278-0062","issn":["0278-0062","1558-254X"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310319808","host_organization_name":"Institute of Electrical and Electronics Engineers","host_organization_lineage":["https://openalex.org/P4310319808"],"host_organization_lineage_names":["Institute of Electrical and Electronics Engineers"],"type":"journal"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"IEEE Transactions on Medical Imaging","raw_type":"journal-article"},{"id":"pmid:33351758","is_oa":false,"landing_page_url":"https://pubmed.ncbi.nlm.nih.gov/33351758","pdf_url":null,"source":{"id":"https://openalex.org/S4306525036","display_name":"PubMed","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I1299303238","host_organization_name":"National Institutes of Health","host_organization_lineage":["https://openalex.org/I1299303238"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"IEEE transactions on medical imaging","raw_type":null},{"id":"pmh:oai:arXiv.org:2102.10365","is_oa":true,"landing_page_url":"http://arxiv.org/abs/2102.10365","pdf_url":"https://arxiv.org/pdf/2102.10365","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},{"id":"pmh:oai:spiral.imperial.ac.uk:10044/1/86338","is_oa":false,"landing_page_url":"http://hdl.handle.net/10044/1/86338","pdf_url":null,"source":{"id":"https://openalex.org/S4306401396","display_name":"Spiral (Imperial College London)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I47508984","host_organization_name":"Imperial College London","host_organization_lineage":["https://openalex.org/I47508984"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":"1077","raw_type":"Journal Article"},{"id":"pmh:oai:spiral.imperial.ac.uk:10044/1/87699","is_oa":false,"landing_page_url":"http://hdl.handle.net/10044/1/87699","pdf_url":null,"source":{"id":"https://openalex.org/S4306401396","display_name":"Spiral (Imperial College London)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I47508984","host_organization_name":"Imperial College London","host_organization_lineage":["https://openalex.org/I47508984"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":"1077","raw_type":"Journal Article"}],"best_oa_location":{"id":"pmh:oai:arXiv.org:2102.10365","is_oa":true,"landing_page_url":"http://arxiv.org/abs/2102.10365","pdf_url":"https://arxiv.org/pdf/2102.10365","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"sustainable_development_goals":[],"awards":[{"id":"https://openalex.org/G4104391792","display_name":null,"funder_award_id":"757173","funder_id":"https://openalex.org/F4320320300","funder_display_name":"European Commission"}],"funders":[{"id":"https://openalex.org/F4320314731","display_name":"UK Research and Innovation","ror":"https://ror.org/001aqnf71"},{"id":"https://openalex.org/F4320320300","display_name":"European Commission","ror":"https://ror.org/00k4n6c32"},{"id":"https://openalex.org/F4320322725","display_name":"China Scholarship Council","ror":"https://ror.org/04atp4p48"}],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":48,"referenced_works":["https://openalex.org/W1884191083","https://openalex.org/W2040010062","https://openalex.org/W2142261479","https://openalex.org/W2148143831","https://openalex.org/W2301358467","https://openalex.org/W2322371438","https://openalex.org/W2342591535","https://openalex.org/W2440599146","https://openalex.org/W2551429935","https://openalex.org/W2592929672","https://openalex.org/W2767106145","https://openalex.org/W2784163702","https://openalex.org/W2798869704","https://openalex.org/W2884821079","https://openalex.org/W2891155035","https://openalex.org/W2903617078","https://openalex.org/W2913592748","https://openalex.org/W2936503027","https://openalex.org/W2949736877","https://openalex.org/W2962767316","https://openalex.org/W2962914239","https://openalex.org/W2962933664","https://openalex.org/W2963351448","https://openalex.org/W2963691377","https://openalex.org/W2964159205","https://openalex.org/W2969985801","https://openalex.org/W2979849298","https://openalex.org/W2992308087","https://openalex.org/W3027849539","https://openalex.org/W3034601242","https://openalex.org/W6631561801","https://openalex.org/W6640425456","https://openalex.org/W6674330103","https://openalex.org/W6681151457","https://openalex.org/W6735013348","https://openalex.org/W6744066916","https://openalex.org/W6744169265","https://openalex.org/W6745136726","https://openalex.org/W6749556305","https://openalex.org/W6752314297","https://openalex.org/W6761292208","https://openalex.org/W6764733053","https://openalex.org/W6766318054","https://openalex.org/W6768920361","https://openalex.org/W6771245373","https://openalex.org/W6771378132","https://openalex.org/W6783399553","https://openalex.org/W6910555300"],"related_works":[],"abstract_inverted_index":{"Class":[0],"imbalance":[1,54],"poses":[2],"a":[3,99,134],"challenge":[4],"for":[5,110],"developing":[6],"unbiased,":[7],"accurate":[8],"predictive":[9],"models.":[10],"In":[11,40],"particular,":[12],"in":[13,32],"image":[14],"segmentation":[15,165,183],"neural":[16],"networks":[17],"may":[18,81],"overfit":[19],"to":[20,37,98,150,174,180,186],"the":[21,33,48,57,76,84,90,155,171,175],"foreground":[22],"samples":[23,88],"from":[24],"small":[25,103],"structures,":[26],"which":[27,146],"are":[28,147,160],"often":[29],"heavily":[30],"under-represented":[31,156],"training":[34,65],"set,":[35],"leading":[36],"poor":[38],"generalization.":[39],"this":[41,119],"study,":[42],"we":[43,121],"provide":[44],"new":[45,123],"insights":[46],"on":[47,162],"problem":[49],"of":[50,78,89,102,126,154],"overfitting":[51],"under":[52],"class":[53,71,92],"by":[55],"inspecting":[56],"network":[58,115],"behavior.":[59],"We":[60],"find":[61],"empirically":[62],"that":[63,170],"when":[64],"with":[66],"limited":[67],"data":[68,144],"and":[69,114,130,143,188],"strong":[70],"imbalance,":[72],"at":[73],"test":[74],"time":[75],"distribution":[77],"logit":[79,152],"activations":[80],"shift":[82,153],"across":[83],"decision":[85],"boundary,":[86],"while":[87],"well-represented":[91],"seem":[93],"unaffected.":[94],"This":[95,105],"bias":[96],"leads":[97],"systematic":[100],"under-segmentation":[101],"structures.":[104],"phenomenon":[106],"is":[107],"consistently":[108],"observed":[109],"different":[111],"databases,":[112],"tasks":[113],"architectures.":[116],"To":[117],"tackle":[118],"problem,":[120],"introduce":[122],"asymmetric":[124],"variants":[125],"popular":[127],"loss":[128],"functions":[129],"regularization":[131],"techniques":[132],"including":[133],"large":[135],"margin":[136],"loss,":[137,139],"focal":[138],"adversarial":[140],"training,":[141],"mixup":[142],"augmentation,":[145],"explicitly":[148],"designed":[149],"counter":[151],"classes.":[157],"Extensive":[158],"experiments":[159],"conducted":[161],"several":[163],"challenging":[164],"tasks.":[166],"Our":[167],"results":[168],"demonstrate":[169],"proposed":[172],"modifications":[173],"objective":[176],"function":[177],"can":[178],"lead":[179],"significantly":[181],"improved":[182],"accuracy":[184],"compared":[185],"baselines":[187],"alternative":[189],"approaches.":[190]},"counts_by_year":[{"year":2026,"cited_by_count":11},{"year":2025,"cited_by_count":51},{"year":2024,"cited_by_count":56},{"year":2023,"cited_by_count":37},{"year":2022,"cited_by_count":28},{"year":2021,"cited_by_count":5}],"updated_date":"2026-07-13T07:31:44.756512","created_date":"2021-01-05T00:00:00"}