{"id":"https://openalex.org/W4399657271","doi":"https://doi.org/10.48550/arxiv.2406.07844","title":"Improving Compositional Attribute Binding in Text-to-Image Generative Models via Enhanced Text Embeddings","display_name":"Improving Compositional Attribute Binding in Text-to-Image Generative Models via Enhanced Text Embeddings","publication_year":2024,"publication_date":"2024-06-12","ids":{"openalex":"https://openalex.org/W4399657271","doi":"https://doi.org/10.48550/arxiv.2406.07844"},"language":"en","primary_location":{"id":"pmh:oai:arXiv.org:2406.07844","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2406.07844","pdf_url":"https://arxiv.org/pdf/2406.07844","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,"listed_in":[],"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":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"volume":null,"type":"preprint","indexed_in":["arxiv","datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://arxiv.org/pdf/2406.07844","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":null,"display_name":"Zarei, Arman","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zarei, Arman","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5019127827","display_name":"Keivan Rezaei","orcid":"https://orcid.org/0000-0003-4295-0832","observed_orcids":["https://orcid.org/0000-0003-4295-0832"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Rezaei, Keivan","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5085795724","display_name":"Samyadeep Basu","orcid":"https://orcid.org/0009-0007-4021-0239","observed_orcids":["https://orcid.org/0009-0007-4021-0239"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Basu, Samyadeep","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5046938640","display_name":"Mehrdad Saberi","orcid":"https://orcid.org/0009-0009-2095-4219","observed_orcids":["https://orcid.org/0009-0009-2095-4219"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Saberi, Mehrdad","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5077078550","display_name":"Mazda Moayeri","orcid":"https://orcid.org/0009-0006-2842-5352","observed_orcids":["https://orcid.org/0009-0006-2842-5352"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Moayeri, Mazda","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5028601879","display_name":"Priyatham Kattakinda","orcid":"https://orcid.org/0000-0002-5179-6585","observed_orcids":["https://orcid.org/0000-0002-5179-6585"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Kattakinda, Priyatham","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5025450606","display_name":"Soheil Feizi","orcid":"https://orcid.org/0000-0003-0944-8242","observed_orcids":["https://orcid.org/0000-0003-0944-8242"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Feizi, Soheil","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":0.9994999766349792,"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":0.9994999766349792,"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":0.00019999999494757503,"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":0.0,"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/text-to-image-generation","display_name":"text-to-image generation","score":0.8479999899864197},{"id":"https://openalex.org/keywords/clip-text-encoder","display_name":"CLIP text encoder","score":0.8090000152587891},{"id":"https://openalex.org/keywords/text-embeddings","display_name":"text embeddings","score":0.8619999885559082},{"id":"https://openalex.org/keywords/diffusion-models","display_name":"diffusion models","score":0.6819999814033508},{"id":"https://openalex.org/keywords/stable-diffusion","display_name":"Stable Diffusion","score":0.9150000214576721},{"id":"https://openalex.org/keywords/prompt-conditioning","display_name":"prompt conditioning","score":0.3970000147819519}],"concepts":[{"id":"https://openalex.org/C39890363","wikidata":"https://www.wikidata.org/wiki/Q36108","display_name":"Generative grammar","level":2,"score":0.7284126281738281},{"id":"https://openalex.org/C115961682","wikidata":"https://www.wikidata.org/wiki/Q860623","display_name":"Image (mathematics)","level":2,"score":0.5520341396331787},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.4895258843898773},{"id":"https://openalex.org/C167966045","wikidata":"https://www.wikidata.org/wiki/Q5532625","display_name":"Generative model","level":3,"score":0.4584197402000427},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.32955098152160645},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.3038417398929596}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:arXiv.org:2406.07844","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2406.07844","pdf_url":"https://arxiv.org/pdf/2406.07844","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,"listed_in":[],"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":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},{"id":"doi:10.48550/arxiv.2406.07844","doi":"https://doi.org/10.48550/arxiv.2406.07844","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2406.07844","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,"listed_in":[],"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":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":null,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"pmh:oai:arXiv.org:2406.07844","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2406.07844","pdf_url":"https://arxiv.org/pdf/2406.07844","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,"listed_in":[],"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":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"sustainable_development_goals":[],"sustainable_development_goals_aurora":[],"study_designs":[],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":["https://openalex.org/W4365211920","https://openalex.org/W3014948380","https://openalex.org/W3204019825","https://openalex.org/W4380551139","https://openalex.org/W4317695495","https://openalex.org/W4287117424","https://openalex.org/W4387506531","https://openalex.org/W4238433571","https://openalex.org/W3174044702","https://openalex.org/W2967848559"],"abstract_inverted_index":{"Text-to-image":[0],"diffusion-based":[1],"generative":[2,33],"models":[3,34,100],"have":[4],"the":[5,26,58,73,77,92,96,127,131,138,171],"stunning":[6],"ability":[7],"to":[8,61,101],"generate":[9,102,120],"photo-realistic":[10],"images":[11],"and":[12,40,80,136,180],"achieve":[13],"state-of-the-art":[14],"low":[15],"FID":[16,173],"scores":[17],"on":[18,184],"challenging":[19],"image":[20],"generation":[21],"benchmarks.":[22],"However,":[23],"one":[24,90],"of":[25,30,91,98,130,195],"primary":[27,93],"failure":[28],"modes":[29],"these":[31,99],"text-to-image":[32],"is":[35,89,134],"in":[36,76,142,155,188],"composing":[37],"attributes,":[38],"objects,":[39],"their":[41],"associated":[42],"relationships":[43],"accurately":[44],"into":[45],"an":[46,114],"image.":[47],"In":[48,106],"our":[49],"paper,":[50],"we":[51,108],"investigate":[52],"compositional":[53,104,123,156,164,196],"attribute":[54],"binding":[55],"failures,":[56],"where":[57],"model":[59],"fails":[60],"correctly":[62],"associate":[63],"descriptive":[64],"attributes":[65],"(such":[66],"as":[67,146],"color,":[68],"shape,":[69],"or":[70],"texture)":[71],"with":[72,86],"corresponding":[74],"objects":[75],"generated":[78],"images,":[79],"highlight":[81],"that":[82,110,118,126,162],"imperfect":[83],"text":[84],"conditioning":[85],"CLIP":[87,132,143],"text-encoder":[88,133],"reasons":[94],"behind":[95],"inability":[97],"high-fidelity":[103],"scenes.":[105],"particular,":[107],"show":[109],"(i)":[111],"there":[112],"exists":[113],"optimal":[115],"text-embedding":[116],"space":[117,129,187],"can":[119,166],"highly":[121],"coherent":[122],"scenes":[124],"showing":[125],"output":[128],"sub-optimal,":[135],"(ii)":[137],"final":[139],"token":[140],"embeddings":[141],"are":[144],"erroneous":[145],"they":[147],"often":[148],"include":[149],"attention":[150],"contributions":[151],"from":[152],"unrelated":[153],"tokens":[154],"prompts.":[157],"Our":[158],"main":[159],"finding":[160],"shows":[161],"significant":[163],"improvements":[165],"be":[167],"achieved":[168],"(without":[169],"harming":[170],"model's":[172],"score)":[174],"by":[175],"fine-tuning":[176],"only":[177],"a":[178,192],"simple":[179],"parameter-efficient":[181],"linear":[182],"projection":[183],"CLIP's":[185],"representation":[186],"Stable-Diffusion":[189],"variants":[190],"using":[191],"small":[193],"set":[194],"image-text":[197],"pairs.":[198]},"counts_by_year":[],"updated_date":"2026-09-29T14:09:16.559479","created_date":"2024-06-14T00:00:00"}