{"id":"https://openalex.org/W7155386784","doi":"https://doi.org/10.48550/arxiv.2604.20258","title":"Rethinking Where to Edit: Task-Aware Localization for Instruction-Based Image Editing","display_name":"Rethinking Where to Edit: Task-Aware Localization for Instruction-Based Image Editing","publication_year":2026,"publication_date":"2026-04-22","ids":{"openalex":"https://openalex.org/W7155386784","doi":"https://doi.org/10.48550/arxiv.2604.20258"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2604.20258","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2604.20258","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"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.2604.20258","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5101532408","display_name":"Jingxuan He","orcid":"https://orcid.org/0009-0003-9858-5640"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"He, Jingxuan","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5134401656","display_name":"Xiyu Wang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wang, Xiyu","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5134425024","display_name":"Mengyu Zheng","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zheng, Mengyu","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5134437787","display_name":"Xiangyu Zeng","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zeng, Xiangyu","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5134449364","display_name":"Yunke Wang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wang, Yunke","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5134371030","display_name":"Chang Xu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Xu, Chang","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":0.705299973487854,"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":0.705299973487854,"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/T12859","display_name":"Cell Image Analysis Techniques","score":0.030700000002980232,"subfield":{"id":"https://openalex.org/subfields/1304","display_name":"Biophysics"},"field":{"id":"https://openalex.org/fields/13","display_name":"Biochemistry, Genetics and Molecular Biology"},"domain":{"id":"https://openalex.org/domains/1","display_name":"Life Sciences"}},{"id":"https://openalex.org/T12357","display_name":"Digital Media Forensic Detection","score":0.015799999237060547,"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/image-editing","display_name":"Image editing","score":0.781000018119812},{"id":"https://openalex.org/keywords/image","display_name":"Image (mathematics)","score":0.5950000286102295},{"id":"https://openalex.org/keywords/construct","display_name":"Construct (python library)","score":0.5273000001907349},{"id":"https://openalex.org/keywords/feature","display_name":"Feature (linguistics)","score":0.4984999895095825},{"id":"https://openalex.org/keywords/consistency","display_name":"Consistency (knowledge bases)","score":0.47429999709129333},{"id":"https://openalex.org/keywords/exploit","display_name":"Exploit","score":0.4684999883174896},{"id":"https://openalex.org/keywords/partition","display_name":"Partition (number theory)","score":0.4090000092983246},{"id":"https://openalex.org/keywords/centroid","display_name":"Centroid","score":0.33410000801086426}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8402000069618225},{"id":"https://openalex.org/C2776674983","wikidata":"https://www.wikidata.org/wiki/Q545981","display_name":"Image editing","level":3,"score":0.781000018119812},{"id":"https://openalex.org/C115961682","wikidata":"https://www.wikidata.org/wiki/Q860623","display_name":"Image (mathematics)","level":2,"score":0.5950000286102295},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5802000164985657},{"id":"https://openalex.org/C2780801425","wikidata":"https://www.wikidata.org/wiki/Q5164392","display_name":"Construct (python library)","level":2,"score":0.5273000001907349},{"id":"https://openalex.org/C2776401178","wikidata":"https://www.wikidata.org/wiki/Q12050496","display_name":"Feature (linguistics)","level":2,"score":0.4984999895095825},{"id":"https://openalex.org/C31972630","wikidata":"https://www.wikidata.org/wiki/Q844240","display_name":"Computer vision","level":1,"score":0.4787999987602234},{"id":"https://openalex.org/C2776436953","wikidata":"https://www.wikidata.org/wiki/Q5163215","display_name":"Consistency (knowledge bases)","level":2,"score":0.47429999709129333},{"id":"https://openalex.org/C165696696","wikidata":"https://www.wikidata.org/wiki/Q11287","display_name":"Exploit","level":2,"score":0.4684999883174896},{"id":"https://openalex.org/C42812","wikidata":"https://www.wikidata.org/wiki/Q1082910","display_name":"Partition (number theory)","level":2,"score":0.4090000092983246},{"id":"https://openalex.org/C146599234","wikidata":"https://www.wikidata.org/wiki/Q511093","display_name":"Centroid","level":2,"score":0.33410000801086426},{"id":"https://openalex.org/C52622490","wikidata":"https://www.wikidata.org/wiki/Q1026626","display_name":"Feature extraction","level":2,"score":0.3188999891281128},{"id":"https://openalex.org/C153180895","wikidata":"https://www.wikidata.org/wiki/Q7148389","display_name":"Pattern recognition (psychology)","level":2,"score":0.3158000111579895},{"id":"https://openalex.org/C9417928","wikidata":"https://www.wikidata.org/wiki/Q1070689","display_name":"Image processing","level":3,"score":0.314300000667572},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.30640000104904175},{"id":"https://openalex.org/C44359876","wikidata":"https://www.wikidata.org/wiki/Q5338467","display_name":"Edit distance","level":2,"score":0.30149999260902405},{"id":"https://openalex.org/C2781238097","wikidata":"https://www.wikidata.org/wiki/Q175026","display_name":"Object (grammar)","level":2,"score":0.2815999984741211},{"id":"https://openalex.org/C2780967703","wikidata":"https://www.wikidata.org/wiki/Q2571389","display_name":"Collaborative editing","level":2,"score":0.272599995136261},{"id":"https://openalex.org/C124504099","wikidata":"https://www.wikidata.org/wiki/Q56933","display_name":"Image segmentation","level":3,"score":0.2669999897480011}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2604.20258","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2604.20258","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"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.2604.20258","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2604.20258","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"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":[{"display_name":"Quality Education","id":"https://metadata.un.org/sdg/4","score":0.5610077381134033}],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Instruction-based":[0],"image":[1,98,105,155,194],"editing":[2,57,159,195],"(IIE)":[3],"aims":[4],"to":[5,9,31,34,123],"modify":[6],"images":[7],"according":[8],"textual":[10],"instructions":[11],"while":[12,184],"preserving":[13],"irrelevant":[14],"content.":[15],"Despite":[16],"recent":[17,193],"advances":[18],"in":[19,75],"diffusion":[20],"transformers,":[21],"existing":[22],"methods":[23],"often":[24],"suffer":[25],"from":[26,44],"over-editing,":[27],"introducing":[28],"unintended":[29],"changes":[30],"regions":[32],"unrelated":[33],"the":[35,45,93,133],"desired":[36],"edit.":[37],"We":[38,107,161],"identify":[39],"that":[40,91,135,149],"this":[41,81],"limitation":[42],"arises":[43],"lack":[46],"of":[47,191],"an":[48],"explicit":[49],"mechanism":[50],"for":[51,157,166],"edit":[52,88,111,127],"localization.":[53],"In":[54],"particular,":[55],"different":[56,158],"operations":[58],"(e.g.,":[59],"addition,":[60],"removal":[61],"and":[62,96,113,128,153,170,199],"replacement)":[63],"induce":[64],"distinct":[65],"spatial":[66],"patterns,":[67],"yet":[68],"current":[69],"IIE":[70,101],"models":[71],"typically":[72],"treat":[73],"localization":[74,89,137],"a":[76,85,144,163],"task-agnostic":[77],"manner.":[78],"To":[79],"address":[80],"limitation,":[82],"we":[83,141],"propose":[84],"training-free,":[86],"task-aware":[87],"framework":[90,178],"exploits":[92],"intrinsic":[94],"source":[95,152],"target":[97,154],"streams":[99,156],"within":[100],"models.":[102],"For":[103],"each":[104],"stream,":[106],"first":[108],"obtain":[109],"attention-based":[110],"cues,":[112],"then":[114],"construct":[115],"feature":[116],"centroids":[117],"based":[118],"on":[119,132,174,189],"these":[120],"attentive":[121],"cues":[122],"partition":[124],"tokens":[125],"into":[126],"non-edit":[129,181],"regions.":[130],"Based":[131],"observation":[134],"optimal":[136],"is":[138],"inherently":[139],"task-dependent,":[140],"further":[142],"introduce":[143],"unified":[145],"mask":[146],"construction":[147],"strategy":[148],"selectively":[150],"leverages":[151],"tasks.":[160],"provide":[162],"systematic":[164],"analysis":[165],"our":[167,177],"proposed":[168],"insights":[169],"approaches.":[171],"Extensive":[172],"experiments":[173],"EdiVal-Bench":[175],"demonstrate":[176],"consistently":[179],"improves":[180],"region":[182],"consistency":[183],"maintaining":[185],"strong":[186],"instruction-following":[187],"performance":[188],"top":[190],"powerful":[192],"backbones,":[196],"including":[197],"Step1X-Edit":[198],"Qwen-Image-Edit.":[200]},"counts_by_year":[],"updated_date":"2026-07-28T07:46:37.118299","created_date":"2026-04-24T00:00:00"}