{"id":"https://openalex.org/W7135039738","doi":"https://doi.org/10.48550/arxiv.2603.10210","title":"Delta-K: Boosting Multi-Instance Generation via Cross-Attention Augmentation","display_name":"Delta-K: Boosting Multi-Instance Generation via Cross-Attention Augmentation","publication_year":2026,"publication_date":"2026-03-10","ids":{"openalex":"https://openalex.org/W7135039738","doi":"https://doi.org/10.48550/arxiv.2603.10210"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2603.10210","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2603.10210","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.2603.10210","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5128873331","display_name":"Zitong Wang","orcid":null},"institutions":[],"countries":[],"is_corresponding":true,"raw_author_name":"Wang, Zitong","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5125691315","display_name":"Zijun Shen","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Shen, Zijun","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5125971926","display_name":"Haohao Xu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Xu, Haohao","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5110221104","display_name":"Z. Luo","orcid":"https://orcid.org/0009-0003-0224-8827"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Luo, Zhengjie","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5128850284","display_name":"Weibin Wu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wu, Weibin","raw_affiliation_strings":[],"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":5,"corresponding_author_ids":["https://openalex.org/A5128873331"],"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":0.47530001401901245,"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":0.47530001401901245,"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/T11714","display_name":"Multimodal Machine Learning Applications","score":0.07079999893903732,"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/T10036","display_name":"Advanced Neural Network Applications","score":0.054499998688697815,"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/inference","display_name":"Inference","score":0.6642000079154968},{"id":"https://openalex.org/keywords/generality","display_name":"Generality","score":0.5900999903678894},{"id":"https://openalex.org/keywords/boosting","display_name":"Boosting (machine learning)","score":0.5771999955177307},{"id":"https://openalex.org/keywords/noise","display_name":"Noise (video)","score":0.459199994802475},{"id":"https://openalex.org/keywords/key","display_name":"Key (lock)","score":0.41589999198913574},{"id":"https://openalex.org/keywords/scheduling","display_name":"Scheduling (production processes)","score":0.3862000107765198},{"id":"https://openalex.org/keywords/semantics","display_name":"Semantics (computer science)","score":0.37619999051094055}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7616999745368958},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.6642000079154968},{"id":"https://openalex.org/C2780767217","wikidata":"https://www.wikidata.org/wiki/Q5532421","display_name":"Generality","level":2,"score":0.5900999903678894},{"id":"https://openalex.org/C46686674","wikidata":"https://www.wikidata.org/wiki/Q466303","display_name":"Boosting (machine learning)","level":2,"score":0.5771999955177307},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5720999836921692},{"id":"https://openalex.org/C99498987","wikidata":"https://www.wikidata.org/wiki/Q2210247","display_name":"Noise (video)","level":3,"score":0.459199994802475},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.41589999198913574},{"id":"https://openalex.org/C206729178","wikidata":"https://www.wikidata.org/wiki/Q2271896","display_name":"Scheduling (production processes)","level":2,"score":0.3862000107765198},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.3856000006198883},{"id":"https://openalex.org/C184337299","wikidata":"https://www.wikidata.org/wiki/Q1437428","display_name":"Semantics (computer science)","level":2,"score":0.37619999051094055},{"id":"https://openalex.org/C63479239","wikidata":"https://www.wikidata.org/wiki/Q7353546","display_name":"Robustness (evolution)","level":3,"score":0.3249000012874603},{"id":"https://openalex.org/C95623464","wikidata":"https://www.wikidata.org/wiki/Q1096149","display_name":"Classifier (UML)","level":2,"score":0.32030001282691956},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.3125999867916107},{"id":"https://openalex.org/C2779696439","wikidata":"https://www.wikidata.org/wiki/Q7512811","display_name":"Signature (topology)","level":2,"score":0.3077999949455261},{"id":"https://openalex.org/C80444323","wikidata":"https://www.wikidata.org/wiki/Q2878974","display_name":"Theoretical computer science","level":1,"score":0.29490000009536743},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.29429998993873596},{"id":"https://openalex.org/C153180895","wikidata":"https://www.wikidata.org/wiki/Q7148389","display_name":"Pattern recognition (psychology)","level":2,"score":0.26350000500679016},{"id":"https://openalex.org/C64876066","wikidata":"https://www.wikidata.org/wiki/Q5141226","display_name":"Cognitive neuroscience of visual object recognition","level":3,"score":0.2628999948501587},{"id":"https://openalex.org/C104267543","wikidata":"https://www.wikidata.org/wiki/Q208163","display_name":"Signal processing","level":3,"score":0.25679999589920044},{"id":"https://openalex.org/C61797465","wikidata":"https://www.wikidata.org/wiki/Q1188986","display_name":"Term (time)","level":2,"score":0.25429999828338623},{"id":"https://openalex.org/C9357733","wikidata":"https://www.wikidata.org/wiki/Q6878417","display_name":"Missing data","level":2,"score":0.25380000472068787}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2603.10210","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2603.10210","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.2603.10210","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2603.10210","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":[{"id":"https://metadata.un.org/sdg/11","display_name":"Sustainable cities and communities","score":0.5298520922660828}],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"While":[0],"Diffusion":[1],"Models":[2],"excel":[3],"in":[4,57],"text-to-image":[5],"synthesis,":[6],"they":[7],"often":[8],"suffer":[9],"from":[10],"concept":[11],"omission":[12,53],"when":[13],"synthesizing":[14],"complex":[15],"multi-instance":[16],"scenes.":[17],"Existing":[18],"training-free":[19],"methods":[20],"attempt":[21],"to":[22],"resolve":[23],"this":[24],"by":[25,54,97],"rescaling":[26],"attention":[27],"maps,":[28],"which":[29],"merely":[30],"exacerbates":[31],"unstructured":[32],"noise":[33,106],"without":[34,137],"establishing":[35],"coherent":[36],"semantic":[37,76,89],"representations.":[38],"To":[39],"address":[40],"this,":[41],"we":[42,67],"propose":[43],"Delta-K,":[44],"a":[45,69,98],"backbone-agnostic":[46],"and":[47,133],"plug-and-play":[48],"inference":[49],"framework":[50],"that":[51,73],"tackles":[52],"operating":[55],"directly":[56],"the":[58,75,87,93,118],"shared":[59],"cross-attention":[60],"Key":[61],"space.":[62],"Specifically,":[63],"with":[64],"Vision-language":[65],"model,":[66],"extract":[68],"differential":[70],"key":[71],"$\u0394K$":[72],"encodes":[74],"signature":[77],"of":[78,92,120],"missing":[79],"concepts.":[80,114],"This":[81],"signal":[82],"is":[83],"then":[84],"injected":[85],"during":[86],"early":[88],"planning":[90],"stage":[91],"diffusion":[94],"process.":[95],"Governed":[96],"dynamically":[99],"optimized":[100],"scheduling":[101],"mechanism,":[102],"Delta-K":[103,123],"grounds":[104],"diffuse":[105],"into":[107],"stable":[108],"structural":[109],"anchors":[110],"while":[111],"preserving":[112],"existing":[113],"Extensive":[115],"experiments":[116],"demonstrate":[117],"generality":[119],"our":[121],"approach:":[122],"consistently":[124],"improves":[125],"compositional":[126],"alignment":[127],"across":[128],"both":[129],"modern":[130],"DiT":[131],"models":[132],"classical":[134],"U-Net":[135],"architectures,":[136],"requiring":[138],"spatial":[139],"masks,":[140],"additional":[141],"training,":[142],"or":[143],"architectural":[144],"modifications.":[145]},"counts_by_year":[],"updated_date":"2026-03-13T14:25:03.468858","created_date":"2026-03-13T00:00:00"}