{"id":"https://openalex.org/W7154153179","doi":"https://doi.org/10.48550/arxiv.2604.09386","title":"Region-Constrained Group Relative Policy Optimization for Flow-Based Image Editing","display_name":"Region-Constrained Group Relative Policy Optimization for Flow-Based Image Editing","publication_year":2026,"publication_date":"2026-04-10","ids":{"openalex":"https://openalex.org/W7154153179","doi":"https://doi.org/10.48550/arxiv.2604.09386"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2604.09386","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2604.09386","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":"article"},"type":"preprint","indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://doi.org/10.48550/arxiv.2604.09386","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5133491219","display_name":"Zhuohan Ouyang","orcid":null},"institutions":[],"countries":[],"is_corresponding":true,"raw_author_name":"Ouyang, Zhuohan","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5133550714","display_name":"Zhe Qian","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Qian, Zhe","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5133483292","display_name":"Wenhuo Cui","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Cui, Wenhuo","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5133496960","display_name":"Chaoqun Wang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wang, Chaoqun","raw_affiliation_strings":[],"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":4,"corresponding_author_ids":["https://openalex.org/A5133491219"],"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":0.46389999985694885,"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":0.46389999985694885,"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":0.09040000289678574,"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/T11019","display_name":"Image Enhancement Techniques","score":0.06599999964237213,"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":0.8392999768257141},{"id":"https://openalex.org/keywords/image","display_name":"Image (mathematics)","score":0.6861000061035156},{"id":"https://openalex.org/keywords/variance","display_name":"Variance (accounting)","score":0.6725999712944031},{"id":"https://openalex.org/keywords/noise","display_name":"Noise (video)","score":0.6097999811172485},{"id":"https://openalex.org/keywords/fidelity","display_name":"Fidelity","score":0.6090999841690063},{"id":"https://openalex.org/keywords/ode","display_name":"Ode","score":0.46470001339912415},{"id":"https://openalex.org/keywords/video-editing","display_name":"Video editing","score":0.4625999927520752},{"id":"https://openalex.org/keywords/high-fidelity","display_name":"High fidelity","score":0.4433000087738037}],"concepts":[{"id":"https://openalex.org/C2776674983","wikidata":"https://www.wikidata.org/wiki/Q545981","display_name":"Image editing","level":3,"score":0.8392999768257141},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7421000003814697},{"id":"https://openalex.org/C115961682","wikidata":"https://www.wikidata.org/wiki/Q860623","display_name":"Image (mathematics)","level":2,"score":0.6861000061035156},{"id":"https://openalex.org/C196083921","wikidata":"https://www.wikidata.org/wiki/Q7915758","display_name":"Variance (accounting)","level":2,"score":0.6725999712944031},{"id":"https://openalex.org/C99498987","wikidata":"https://www.wikidata.org/wiki/Q2210247","display_name":"Noise (video)","level":3,"score":0.6097999811172485},{"id":"https://openalex.org/C2776459999","wikidata":"https://www.wikidata.org/wiki/Q2119376","display_name":"Fidelity","level":2,"score":0.6090999841690063},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.47510001063346863},{"id":"https://openalex.org/C34862557","wikidata":"https://www.wikidata.org/wiki/Q178985","display_name":"Ode","level":2,"score":0.46470001339912415},{"id":"https://openalex.org/C2780310081","wikidata":"https://www.wikidata.org/wiki/Q1154312","display_name":"Video editing","level":2,"score":0.4625999927520752},{"id":"https://openalex.org/C113364801","wikidata":"https://www.wikidata.org/wiki/Q26674","display_name":"High fidelity","level":2,"score":0.4433000087738037},{"id":"https://openalex.org/C2780967703","wikidata":"https://www.wikidata.org/wiki/Q2571389","display_name":"Collaborative editing","level":2,"score":0.43050000071525574},{"id":"https://openalex.org/C160633673","wikidata":"https://www.wikidata.org/wiki/Q355198","display_name":"Pixel","level":2,"score":0.4041999876499176},{"id":"https://openalex.org/C31972630","wikidata":"https://www.wikidata.org/wiki/Q844240","display_name":"Computer vision","level":1,"score":0.3718000054359436},{"id":"https://openalex.org/C2776889888","wikidata":"https://www.wikidata.org/wiki/Q1135789","display_name":"Unintended consequences","level":2,"score":0.3587000072002411},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.3531000018119812},{"id":"https://openalex.org/C160920958","wikidata":"https://www.wikidata.org/wiki/Q7662746","display_name":"Synthetic data","level":2,"score":0.29989999532699585},{"id":"https://openalex.org/C107457646","wikidata":"https://www.wikidata.org/wiki/Q207434","display_name":"Human\u2013computer interaction","level":1,"score":0.2897000014781952},{"id":"https://openalex.org/C2781170535","wikidata":"https://www.wikidata.org/wiki/Q30587856","display_name":"Noisy data","level":2,"score":0.27619999647140503},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.26919999718666077},{"id":"https://openalex.org/C35772409","wikidata":"https://www.wikidata.org/wiki/Q1323086","display_name":"Image noise","level":3,"score":0.26350000500679016},{"id":"https://openalex.org/C80444323","wikidata":"https://www.wikidata.org/wiki/Q2878974","display_name":"Theoretical computer science","level":1,"score":0.2630999982357025},{"id":"https://openalex.org/C9417928","wikidata":"https://www.wikidata.org/wiki/Q1070689","display_name":"Image processing","level":3,"score":0.260699987411499},{"id":"https://openalex.org/C146044194","wikidata":"https://www.wikidata.org/wiki/Q5157334","display_name":"Computational photography","level":4,"score":0.25619998574256897},{"id":"https://openalex.org/C79403827","wikidata":"https://www.wikidata.org/wiki/Q3988","display_name":"Real-time computing","level":1,"score":0.25040000677108765}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2604.09386","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2604.09386","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":"article"}],"best_oa_location":{"id":"doi:10.48550/arxiv.2604.09386","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2604.09386","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":"article"},"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":{"Instruction-guided":[0],"image":[1,24,94],"editing":[2,55,95,112,150,168],"requires":[3],"balancing":[4],"target":[5],"modification":[6],"with":[7,147],"non-target":[8,70,118,159,173],"preservation.":[9,174],"Recently,":[10],"flow-based":[11,93],"models":[12],"have":[13],"emerged":[14],"as":[15],"a":[16,87],"strong":[17],"and":[18,31,54,76,134,138,172],"increasingly":[19],"adopted":[20],"backbone":[21],"for":[22,92],"instruction-guided":[23],"editing,":[25],"thanks":[26],"to":[27,46,105,129],"their":[28],"high":[29],"fidelity":[30],"efficient":[32],"deterministic":[33,97],"ODE":[34,98],"sampling.":[35,99],"Building":[36],"on":[37,162],"this":[38],"foundation,":[39],"GRPO-based":[40],"reward-driven":[41],"post-training":[42,90],"has":[43],"been":[44],"explored":[45],"directly":[47],"optimize":[48],"editing-specific":[49],"rewards,":[50],"improving":[51,111],"instruction":[52,114,170],"following":[53],"consistency.":[56],"However,":[57],"existing":[58],"methods":[59],"often":[60],"suffer":[61],"from":[62],"noisy":[63,78],"credit":[64,109],"assignment:":[65],"global":[66],"exploration":[67,123],"also":[68],"perturbs":[69],"regions,":[71],"inflating":[72],"within-group":[73],"reward":[74,132,143],"variance":[75,104,133],"yielding":[77],"GRPO":[79,89,136],"advantages.":[80],"To":[81],"address":[82],"this,":[83],"we":[84,121],"propose":[85],"RC-GRPO-Editing,":[86],"region-constrained":[88],"framework":[91],"under":[96],"It":[100],"suppresses":[101],"background-induced":[102,131],"nuisance":[103],"enable":[106],"cleaner":[107],"localized":[108],"assignment,":[110],"region":[113,151,169],"adherence":[115,171],"while":[116],"preserving":[117],"content.":[119],"Concretely,":[120],"localize":[122],"via":[124],"region-decoupled":[125],"initial":[126],"noise":[127],"perturbations":[128],"reduce":[130],"stabilize":[135],"advantages,":[137],"introduce":[139],"an":[140],"attention":[141],"concentration":[142],"that":[144],"aligns":[145],"cross-attention":[146],"the":[148,153],"intended":[149],"throughout":[152],"rollout,":[154],"reducing":[155],"unintended":[156],"changes":[157],"in":[158,167],"regions.":[160],"Experiments":[161],"CompBench":[163],"show":[164],"consistent":[165],"improvements":[166]},"counts_by_year":[],"updated_date":"2026-04-14T06:08:25.285971","created_date":"2026-04-14T00:00:00"}