{"id":"https://openalex.org/W4400580465","doi":"https://doi.org/10.1145/3641519.3657527","title":"Separate-and-Enhance: Compositional Finetuning for Text-to-Image Diffusion Models","display_name":"Separate-and-Enhance: Compositional Finetuning for Text-to-Image Diffusion Models","publication_year":2024,"publication_date":"2024-07-12","ids":{"openalex":"https://openalex.org/W4400580465","doi":"https://doi.org/10.1145/3641519.3657527"},"language":"en","primary_location":{"id":"doi:10.1145/3641519.3657527","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3641519.3657527","pdf_url":null,"source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Special Interest Group on Computer Graphics and Interactive Techniques Conference Conference Papers","raw_type":"proceedings-article"},"type":"conference-paper","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://doi.org/10.1145/3641519.3657527","any_repository_has_fulltext":null},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5076213540","display_name":"Zhipeng Bao","orcid":"https://orcid.org/0009-0008-9898-3741"},"institutions":[{"id":"https://openalex.org/I74973139","display_name":"Carnegie Mellon University","ror":"https://ror.org/05x2bcf33","country_code":"US","type":"education","lineage":["https://openalex.org/I74973139"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Zhipeng Bao","raw_affiliation_strings":["Carnegie Mellon University, United States of America"],"raw_orcid":"https://orcid.org/0009-0008-9898-3741","affiliations":[{"raw_affiliation_string":"Carnegie Mellon University, United States of America","institution_ids":["https://openalex.org/I74973139"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100447637","display_name":"Yijun Li","orcid":"https://orcid.org/0000-0001-7295-8750"},"institutions":[{"id":"https://openalex.org/I1306409833","display_name":"Adobe Systems (United States)","ror":"https://ror.org/059tvcg64","country_code":"US","type":"company","lineage":["https://openalex.org/I1306409833"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Yijun Li","raw_affiliation_strings":["Adobe Research, United States of America"],"raw_orcid":"https://orcid.org/0000-0001-7295-8750","affiliations":[{"raw_affiliation_string":"Adobe Research, United States of America","institution_ids":["https://openalex.org/I1306409833"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101960524","display_name":"Krishna Kumar Singh","orcid":"https://orcid.org/0000-0002-8066-6835"},"institutions":[{"id":"https://openalex.org/I1306409833","display_name":"Adobe Systems (United States)","ror":"https://ror.org/059tvcg64","country_code":"US","type":"company","lineage":["https://openalex.org/I1306409833"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Krishna Kumar Singh","raw_affiliation_strings":["Adobe Research, United States of America"],"raw_orcid":"https://orcid.org/0000-0002-8066-6835","affiliations":[{"raw_affiliation_string":"Adobe Research, United States of America","institution_ids":["https://openalex.org/I1306409833"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5049973776","display_name":"Yu-Xiong Wang","orcid":"https://orcid.org/0000-0003-4414-0198"},"institutions":[{"id":"https://openalex.org/I157725225","display_name":"University of Illinois Urbana-Champaign","ror":"https://ror.org/047426m28","country_code":"US","type":"education","lineage":["https://openalex.org/I157725225"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Yu-Xiong Wang","raw_affiliation_strings":["University of Illinois at Urbana-Champaign, United States of America"],"raw_orcid":"https://orcid.org/0000-0003-4414-0198","affiliations":[{"raw_affiliation_string":"University of Illinois at Urbana-Champaign, United States of America","institution_ids":["https://openalex.org/I157725225"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5075246991","display_name":"Martial Hebert","orcid":"https://orcid.org/0000-0003-4566-5930"},"institutions":[{"id":"https://openalex.org/I74973139","display_name":"Carnegie Mellon University","ror":"https://ror.org/05x2bcf33","country_code":"US","type":"education","lineage":["https://openalex.org/I74973139"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Martial Hebert","raw_affiliation_strings":["Carnegie Mellon University, United States of America"],"raw_orcid":"https://orcid.org/0000-0003-4566-5930","affiliations":[{"raw_affiliation_string":"Carnegie Mellon University, United States of America","institution_ids":["https://openalex.org/I74973139"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":3,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":false,"cited_by_count":5,"citation_normalized_percentile":null,"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"1","last_page":"10"},"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":0.977400004863739,"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":0.977400004863739,"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/T12422","display_name":"Radiomics and Machine Learning in Medical Imaging","score":0.9488999843597412,"subfield":{"id":"https://openalex.org/subfields/2741","display_name":"Radiology, Nuclear Medicine and Imaging"},"field":{"id":"https://openalex.org/fields/27","display_name":"Medicine"},"domain":{"id":"https://openalex.org/domains/4","display_name":"Health Sciences"}},{"id":"https://openalex.org/T10824","display_name":"Image Retrieval and Classification Techniques","score":0.9129999876022339,"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/computer-science","display_name":"Computer science","score":0.6480002999305725},{"id":"https://openalex.org/keywords/diffusion","display_name":"Diffusion","score":0.4471403658390045},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.3241971731185913},{"id":"https://openalex.org/keywords/physics","display_name":"Physics","score":0.07782536745071411},{"id":"https://openalex.org/keywords/thermodynamics","display_name":"Thermodynamics","score":0.06677621603012085}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6480002999305725},{"id":"https://openalex.org/C69357855","wikidata":"https://www.wikidata.org/wiki/Q163214","display_name":"Diffusion","level":2,"score":0.4471403658390045},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.3241971731185913},{"id":"https://openalex.org/C121332964","wikidata":"https://www.wikidata.org/wiki/Q413","display_name":"Physics","level":0,"score":0.07782536745071411},{"id":"https://openalex.org/C97355855","wikidata":"https://www.wikidata.org/wiki/Q11473","display_name":"Thermodynamics","level":1,"score":0.06677621603012085}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1145/3641519.3657527","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3641519.3657527","pdf_url":null,"source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Special Interest Group on Computer Graphics and Interactive Techniques Conference Conference Papers","raw_type":"proceedings-article"}],"best_oa_location":{"id":"doi:10.1145/3641519.3657527","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3641519.3657527","pdf_url":null,"source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Special Interest Group on Computer Graphics and Interactive Techniques Conference Conference Papers","raw_type":"proceedings-article"},"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":15,"referenced_works":["https://openalex.org/W2150134853","https://openalex.org/W4206946442","https://openalex.org/W4281485151","https://openalex.org/W4312424618","https://openalex.org/W4312694728","https://openalex.org/W4312824283","https://openalex.org/W4312933868","https://openalex.org/W4386076215","https://openalex.org/W4386076403","https://openalex.org/W4386076425","https://openalex.org/W4390872671","https://openalex.org/W4390873054","https://openalex.org/W4390873084","https://openalex.org/W4390873875","https://openalex.org/W4393154018"],"related_works":["https://openalex.org/W4391375266","https://openalex.org/W2748952813","https://openalex.org/W2390279801","https://openalex.org/W2358668433","https://openalex.org/W4396701345","https://openalex.org/W2376932109","https://openalex.org/W2001405890","https://openalex.org/W4396696052","https://openalex.org/W2382290278","https://openalex.org/W4395014643"],"abstract_inverted_index":{"Despite":[0],"recent":[1],"significant":[2],"strides":[3],"achieved":[4],"by":[5,42],"diffusion-based":[6],"Text-to-Image":[7],"(T2I)":[8],"models,":[9],"current":[10],"systems":[11],"are":[12],"still":[13],"less":[14],"capable":[15],"of":[16,140],"ensuring":[17],"decent":[18],"compositional":[19,57],"generation":[20],"aligned":[21],"with":[22,60,136],"text":[23],"prompts,":[24],"particularly":[25],"for":[26,39],"the":[27,36,64,68,106],"multi-object":[28,102],"generation.":[29],"In":[30],"this":[31],"work,":[32],"we":[33,54,130],"first":[34],"show":[35,131],"fundamental":[37],"reasons":[38],"such":[40],"misalignment":[41],"identifying":[43],"issues":[44],"related":[45],"to":[46,95,144,147,154],"low":[47],"attention":[48,78],"activation":[49],"and":[50,67,76,108,123],"mask":[51,74],"overlaps.":[52],"Then":[53],"propose":[55],"a":[56,137],"finetuning":[58],"framework":[59],"two":[61],"novel":[62,148],"objectives,":[63],"Separate":[65],"loss":[66],"Enhance":[69],"loss,":[70],"that":[71,132],"reduce":[72],"object":[73],"overlaps":[75],"maximize":[77],"scores,":[79],"respectively.":[80],"Unlike":[81],"conventional":[82],"test-time":[83],"adaptation":[84],"methods,":[85],"our":[86,113,134],"model,":[87],"once":[88],"finetuned":[89],"on":[90,157],"critical":[91],"parameters,":[92],"is":[93],"able":[94],"directly":[96],"perform":[97],"inference":[98],"given":[99],"an":[100],"arbitrary":[101],"prompt,":[103],"which":[104],"enhances":[105],"scalability":[107],"generalizability.":[109],"Through":[110],"comprehensive":[111],"evaluations,":[112],"model":[114,135],"demonstrates":[115],"superior":[116],"performance":[117,152],"in":[118],"image":[119],"realism,":[120],"text-image":[121],"alignment,":[122],"adaptability,":[124],"significantly":[125],"surpassing":[126],"established":[127],"baselines.":[128],"Furthermore,":[129],"training":[133],"diverse":[138],"range":[139],"concepts":[141],"enables":[142],"it":[143],"generalize":[145],"effectively":[146],"concepts,":[149],"exhibiting":[150],"enhanced":[151],"compared":[153],"models":[155],"trained":[156],"individual":[158],"concept":[159],"pairs.":[160]},"counts_by_year":[{"year":2026,"cited_by_count":2},{"year":2025,"cited_by_count":3}],"updated_date":"2026-07-14T23:27:15.235271","created_date":"2025-10-10T00:00:00"}