{"id":"https://openalex.org/W7166510311","doi":"https://doi.org/10.48550/arxiv.2606.28226","title":"Exposure Bias Can Alleviate Itself via Directional and Frequency Rectification in Flow Matching","display_name":"Exposure Bias Can Alleviate Itself via Directional and Frequency Rectification in Flow Matching","publication_year":2026,"publication_date":"2026-06-26","ids":{"openalex":"https://openalex.org/W7166510311","doi":"https://doi.org/10.48550/arxiv.2606.28226"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2606.28226","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.28226","pdf_url":null,"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":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"},"type":"preprint","indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://doi.org/10.48550/arxiv.2606.28226","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5111830185","display_name":"Guanbo Huang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Huang, Guanbo","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139611201","display_name":"Jingjia Mao","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Mao, Jingjia","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5113011597","display_name":"Fanding Huang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Huang, Fanding","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139544322","display_name":"Fengkai Liu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Liu, Fengkai","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5090665711","display_name":"Xiaoman Luo","orcid":"https://orcid.org/0000-0001-9987-1399"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Luo, Xiangyang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5107311577","display_name":"Y Liang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Liang, Yaoyuan","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5108898629","display_name":"Jiasheng Lu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Lu, Jiasheng","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139583536","display_name":"Xiaoe Wang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wang, Xiaoe","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139547942","display_name":"Pei Liu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Liu, Pei","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5002602472","display_name":"Ruiliu Fu","orcid":"https://orcid.org/0000-0001-7615-0054"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Fu, Ruiliu","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5113395530","display_name":"Ruqi Huang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Huang, Ruqi","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5043900893","display_name":"Sicong Huang","orcid":"https://orcid.org/0000-0003-0854-4734"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Huang, Shao-Lun","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":0,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":false,"cited_by_count":0,"citation_normalized_percentile":null,"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":null,"last_page":null},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10775","display_name":"Generative Adversarial Networks and Image Synthesis","score":0.33709999918937683,"subfield":{"id":"https://openalex.org/subfields/1707","display_name":"Computer Vision and Pattern Recognition"},"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/T10775","display_name":"Generative Adversarial Networks and Image Synthesis","score":0.33709999918937683,"subfield":{"id":"https://openalex.org/subfields/1707","display_name":"Computer Vision and Pattern Recognition"},"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":0.14350000023841858,"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/T10036","display_name":"Advanced Neural Network Applications","score":0.10580000281333923,"subfield":{"id":"https://openalex.org/subfields/1707","display_name":"Computer Vision and Pattern Recognition"},"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/rectification","display_name":"Rectification","score":0.6650000214576721},{"id":"https://openalex.org/keywords/leverage","display_name":"Leverage (statistics)","score":0.6517000198364258},{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.5975000262260437},{"id":"https://openalex.org/keywords/matching","display_name":"Matching (statistics)","score":0.5059000253677368},{"id":"https://openalex.org/keywords/compensation","display_name":"Compensation (psychology)","score":0.5041000247001648},{"id":"https://openalex.org/keywords/weighting","display_name":"Weighting","score":0.45969998836517334},{"id":"https://openalex.org/keywords/process","display_name":"Process (computing)","score":0.4406999945640564},{"id":"https://openalex.org/keywords/control-theory","display_name":"Control theory (sociology)","score":0.38029998540878296}],"concepts":[{"id":"https://openalex.org/C50942859","wikidata":"https://www.wikidata.org/wiki/Q4967193","display_name":"Rectification","level":3,"score":0.6650000214576721},{"id":"https://openalex.org/C153083717","wikidata":"https://www.wikidata.org/wiki/Q6535263","display_name":"Leverage (statistics)","level":2,"score":0.6517000198364258},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.5975000262260437},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.5916000008583069},{"id":"https://openalex.org/C165064840","wikidata":"https://www.wikidata.org/wiki/Q1321061","display_name":"Matching (statistics)","level":2,"score":0.5059000253677368},{"id":"https://openalex.org/C2780023022","wikidata":"https://www.wikidata.org/wiki/Q1338171","display_name":"Compensation (psychology)","level":2,"score":0.5041000247001648},{"id":"https://openalex.org/C183115368","wikidata":"https://www.wikidata.org/wiki/Q856577","display_name":"Weighting","level":2,"score":0.45969998836517334},{"id":"https://openalex.org/C98045186","wikidata":"https://www.wikidata.org/wiki/Q205663","display_name":"Process (computing)","level":2,"score":0.4406999945640564},{"id":"https://openalex.org/C47446073","wikidata":"https://www.wikidata.org/wiki/Q5165890","display_name":"Control theory (sociology)","level":3,"score":0.38029998540878296},{"id":"https://openalex.org/C70136482","wikidata":"https://www.wikidata.org/wiki/Q13583781","display_name":"A-weighting","level":3,"score":0.376800000667572},{"id":"https://openalex.org/C2779843651","wikidata":"https://www.wikidata.org/wiki/Q7390335","display_name":"SIGNAL (programming language)","level":2,"score":0.3684000074863434},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.3257000148296356},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.2842000126838684},{"id":"https://openalex.org/C97256817","wikidata":"https://www.wikidata.org/wiki/Q1462316","display_name":"Spurious relationship","level":2,"score":0.2705000042915344},{"id":"https://openalex.org/C38349280","wikidata":"https://www.wikidata.org/wiki/Q1434290","display_name":"Flow (mathematics)","level":2,"score":0.26750001311302185},{"id":"https://openalex.org/C185429906","wikidata":"https://www.wikidata.org/wiki/Q1130160","display_name":"Estimator","level":2,"score":0.2671999931335449},{"id":"https://openalex.org/C158600405","wikidata":"https://www.wikidata.org/wiki/Q5054566","display_name":"Causal inference","level":2,"score":0.2644999921321869},{"id":"https://openalex.org/C63479239","wikidata":"https://www.wikidata.org/wiki/Q7353546","display_name":"Robustness (evolution)","level":3,"score":0.2531000077724457},{"id":"https://openalex.org/C167966045","wikidata":"https://www.wikidata.org/wiki/Q5532625","display_name":"Generative model","level":3,"score":0.2515000104904175}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2606.28226","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.28226","pdf_url":null,"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":null,"is_accepted":false,"is_published":null,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"doi:10.48550/arxiv.2606.28226","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.28226","pdf_url":null,"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":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"},"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Flow":[0],"Matching":[1],"(FM)":[2],"has":[3],"achieved":[4],"remarkable":[5],"generative":[6],"performance,":[7],"yet":[8],"it":[9],"suffers":[10],"from":[11,80,140],"exposure":[12,38,71,150],"bias":[13,39,82,88,137,151,159],"due":[14],"to":[15,69,84,107,111,166],"discrepancies":[16],"between":[17],"training":[18,68],"and":[19,76,149,176,184,190],"inference.":[20],"Existing":[21],"mitigation":[22],"strategies":[23],"typically":[24],"rely":[25],"on":[26,173],"static":[27],"constraints":[28],"or":[29],"external":[30],"heuristics.":[31],"In":[32],"this":[33],"work,":[34],"we":[35,54,133],"propose":[36],"that":[37,45,135,179],"itself":[40,83,160],"inherently":[41],"contains":[42],"dynamic":[43],"signals":[44,79],"can":[46],"guide":[47],"its":[48],"own":[49],"rectification.":[50],"To":[51],"leverage":[52],"this,":[53],"introduce":[55],"DEFAR":[56,180],"(DirEctional-Frequency":[57],"Adaptive":[58],"Rectification).":[59],"This":[60],"framework":[61],"simulates":[62],"the":[63,81,86,109,117,121,153,158,168],"single-step":[64],"inference":[65,191],"process":[66],"during":[67],"identify":[70],"bias.":[72],"It":[73,90],"utilizes":[74],"directional":[75],"frequency-adaptive":[77],"feedback":[78],"enhance":[85],"model's":[87],"tolerance.":[89],"consists":[91],"of":[92,143],"two":[93],"key":[94],"components:":[95],"(1)":[96],"Anti-Drift":[97],"Rectification":[98],"(ADR).":[99],"ADR":[100,119],"treats":[101],"inference-time":[102],"drift":[103],"as":[104,161],"a":[105,141,162],"signal":[106],"learn":[108],"direction":[110],"steer":[112],"deviated":[113],"states":[114],"back":[115],"toward":[116],"target.":[118],"endows":[120],"model":[122],"with":[123],"intrinsic":[124],"active":[125],"self-rectification":[126],"capabilities;":[127],"(2)":[128],"Frequency":[129],"Compensation":[130],"(FC).":[131],"Empirically,":[132],"observe":[134],"accumulated":[136],"often":[138],"stems":[139],"lack":[142],"low-frequency":[144],"components":[145],"in":[146],"high-noise":[147],"stages,":[148],"carries":[152],"missing":[154,169],"frequency.":[155],"FC":[156],"leverages":[157],"self-feedback":[163],"weighting":[164],"factor":[165],"reinforce":[167],"frequency":[170],"components.":[171],"Experiments":[172],"CIFAR-10,":[174],"CelebA-64,":[175],"ImageNet-256/512":[177],"show":[178],"outperforms":[181],"prior":[182],"baselines":[183],"further":[185],"demonstrates":[186],"favorable":[187],"scalability,":[188],"compatibility,":[189],"robustness.":[192]},"counts_by_year":[],"updated_date":"2026-07-01T06:00:48.157686","created_date":"2026-06-30T00:00:00"}