{"id":"https://openalex.org/W7131399911","doi":"https://doi.org/10.48550/arxiv.2602.19946","title":"When Pretty Isn't Useful: Investigating Why Modern Text-to-Image Models Fail as Reliable Training Data Generators","display_name":"When Pretty Isn't Useful: Investigating Why Modern Text-to-Image Models Fail as Reliable Training Data Generators","publication_year":2026,"publication_date":"2026-02-23","ids":{"openalex":"https://openalex.org/W7131399911","doi":"https://doi.org/10.48550/arxiv.2602.19946"},"language":null,"primary_location":{"id":"pmh:doi:10.48550/arxiv.2602.19946","is_oa":true,"landing_page_url":null,"pdf_url":null,"source":{"id":"https://openalex.org/S4406922384","display_name":"Open MIND","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"repository"},"license":"publisher-specific-oa","license_id":"https://openalex.org/licenses/publisher-specific-oa","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Article"},"type":"article","indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":null,"any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5056206769","display_name":"Krzysztof Adamkiewicz","orcid":"https://orcid.org/0009-0006-0730-6899"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Adamkiewicz, Krzysztof","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5069729032","display_name":"Brian B. Moser","orcid":"https://orcid.org/0000-0002-0290-7904"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Moser, Brian Bernhard","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5053202923","display_name":"Stanislav Frolov","orcid":"https://orcid.org/0000-0003-2700-5031"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Frolov, Stanislav","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5013376237","display_name":"Tobias Christian Nauen","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Nauen, Tobias Christian","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5029509133","display_name":"Federico Raue","orcid":"https://orcid.org/0000-0002-8604-6207"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Raue, Federico","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":null,"display_name":"Dengel, Andreas","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Dengel, Andreas","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":0.0,"has_fulltext":false,"cited_by_count":0,"citation_normalized_percentile":{"value":0.2201508,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"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":0.38350000977516174,"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":0.38350000977516174,"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/T12650","display_name":"Aesthetic Perception and Analysis","score":0.2303999960422516,"subfield":{"id":"https://openalex.org/subfields/2805","display_name":"Cognitive Neuroscience"},"field":{"id":"https://openalex.org/fields/28","display_name":"Neuroscience"},"domain":{"id":"https://openalex.org/domains/1","display_name":"Life Sciences"}},{"id":"https://openalex.org/T11094","display_name":"Face Recognition and Perception","score":0.06870000064373016,"subfield":{"id":"https://openalex.org/subfields/2805","display_name":"Cognitive Neuroscience"},"field":{"id":"https://openalex.org/fields/28","display_name":"Neuroscience"},"domain":{"id":"https://openalex.org/domains/1","display_name":"Life Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/fidelity","display_name":"Fidelity","score":0.6171000003814697},{"id":"https://openalex.org/keywords/scalability","display_name":"Scalability","score":0.5361999869346619},{"id":"https://openalex.org/keywords/synthetic-data","display_name":"Synthetic data","score":0.5091999769210815},{"id":"https://openalex.org/keywords/training-set","display_name":"Training set","score":0.5080999732017517},{"id":"https://openalex.org/keywords/training","display_name":"Training (meteorology)","score":0.3921000063419342},{"id":"https://openalex.org/keywords/test-data","display_name":"Test data","score":0.38040000200271606},{"id":"https://openalex.org/keywords/generative-model","display_name":"Generative model","score":0.38029998540878296},{"id":"https://openalex.org/keywords/generative-grammar","display_name":"Generative grammar","score":0.366100013256073}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6603000164031982},{"id":"https://openalex.org/C2776459999","wikidata":"https://www.wikidata.org/wiki/Q2119376","display_name":"Fidelity","level":2,"score":0.6171000003814697},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.5482000112533569},{"id":"https://openalex.org/C48044578","wikidata":"https://www.wikidata.org/wiki/Q727490","display_name":"Scalability","level":2,"score":0.5361999869346619},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5321999788284302},{"id":"https://openalex.org/C160920958","wikidata":"https://www.wikidata.org/wiki/Q7662746","display_name":"Synthetic data","level":2,"score":0.5091999769210815},{"id":"https://openalex.org/C51632099","wikidata":"https://www.wikidata.org/wiki/Q3985153","display_name":"Training set","level":2,"score":0.5080999732017517},{"id":"https://openalex.org/C2777211547","wikidata":"https://www.wikidata.org/wiki/Q17141490","display_name":"Training (meteorology)","level":2,"score":0.3921000063419342},{"id":"https://openalex.org/C16910744","wikidata":"https://www.wikidata.org/wiki/Q7705759","display_name":"Test data","level":2,"score":0.38040000200271606},{"id":"https://openalex.org/C167966045","wikidata":"https://www.wikidata.org/wiki/Q5532625","display_name":"Generative model","level":3,"score":0.38029998540878296},{"id":"https://openalex.org/C39890363","wikidata":"https://www.wikidata.org/wiki/Q36108","display_name":"Generative grammar","level":2,"score":0.366100013256073},{"id":"https://openalex.org/C67186912","wikidata":"https://www.wikidata.org/wiki/Q367664","display_name":"Data modeling","level":2,"score":0.35420000553131104},{"id":"https://openalex.org/C113364801","wikidata":"https://www.wikidata.org/wiki/Q26674","display_name":"High fidelity","level":2,"score":0.335099995136261},{"id":"https://openalex.org/C43214815","wikidata":"https://www.wikidata.org/wiki/Q7310987","display_name":"Reliability (semiconductor)","level":3,"score":0.31290000677108765},{"id":"https://openalex.org/C3020493868","wikidata":"https://www.wikidata.org/wiki/Q55631277","display_name":"Real world data","level":2,"score":0.31139999628067017},{"id":"https://openalex.org/C75684735","wikidata":"https://www.wikidata.org/wiki/Q858810","display_name":"Big data","level":2,"score":0.30410000681877136},{"id":"https://openalex.org/C172367668","wikidata":"https://www.wikidata.org/wiki/Q6504956","display_name":"Data visualization","level":3,"score":0.289900004863739},{"id":"https://openalex.org/C2522767166","wikidata":"https://www.wikidata.org/wiki/Q2374463","display_name":"Data science","level":1,"score":0.2892000079154968},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.2791999876499176},{"id":"https://openalex.org/C2777267654","wikidata":"https://www.wikidata.org/wiki/Q3519023","display_name":"Test (biology)","level":2,"score":0.2750000059604645},{"id":"https://openalex.org/C36464697","wikidata":"https://www.wikidata.org/wiki/Q451553","display_name":"Visualization","level":2,"score":0.26350000500679016},{"id":"https://openalex.org/C2780440489","wikidata":"https://www.wikidata.org/wiki/Q5227278","display_name":"Data-driven","level":2,"score":0.2515999972820282}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:doi:10.48550/arxiv.2602.19946","is_oa":true,"landing_page_url":null,"pdf_url":null,"source":{"id":"https://openalex.org/S4406922384","display_name":"Open MIND","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"repository"},"license":"publisher-specific-oa","license_id":"https://openalex.org/licenses/publisher-specific-oa","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Article"},{"id":"doi:10.48550/arxiv.2602.19946","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2602.19946","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":"pmh:doi:10.48550/arxiv.2602.19946","is_oa":true,"landing_page_url":null,"pdf_url":null,"source":{"id":"https://openalex.org/S4406922384","display_name":"Open MIND","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"repository"},"license":"publisher-specific-oa","license_id":"https://openalex.org/licenses/publisher-specific-oa","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Article"},"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Recent":[0],"text-to-image":[1],"(T2I)":[2],"diffusion":[3],"models":[4,56,97,109,158],"produce":[5],"visually":[6],"stunning":[7],"images":[8],"and":[9,42,60,70,83,119],"demonstrate":[10],"excellent":[11],"prompt":[12,84],"following.":[13],"But":[14],"do":[15],"they":[16],"perform":[17],"well":[18],"as":[19,34,98,159],"synthetic":[20,32,51,68],"vision":[21,132],"data":[22,33,91,100,121,143,162],"generators?":[23],"In":[24],"this":[25,67],"work,":[26],"we":[27],"revisit":[28],"the":[29,153],"promise":[30],"of":[31,155],"a":[35,44,105,112,128],"scalable":[36],"substitute":[37],"for":[38],"real":[39,74,89,120],"training":[40,99,161],"sets":[41],"uncover":[43],"surprising":[45],"performance":[46],"regression.":[47],"We":[48,145],"generate":[49],"large-scale":[50],"datasets":[52],"using":[53],"state-of-the-art":[54],"T2I":[55,96,157],"released":[57],"between":[58],"2022":[59],"2025,":[61],"train":[62],"standard":[63],"classifiers":[64],"solely":[65],"on":[66,73,88],"data,":[69],"evaluate":[71],"them":[72],"test":[75,90],"data.":[76],"Despite":[77],"observable":[78],"advances":[79],"in":[80,131,137,142],"visual":[81],"fidelity":[82],"adherence,":[85],"classification":[86],"accuracy":[87],"consistently":[92],"declines":[93],"with":[94],"newer":[95],"generators.":[101,163],"Our":[102],"analysis":[103],"reveals":[104],"hidden":[106],"trend:":[107],"These":[108],"collapse":[110],"to":[111,151],"narrow,":[113],"aesthetic-centric":[114],"distribution":[115,122],"that":[116,135],"undermines":[117],"diversity":[118],"coverage.":[123],"Overall,":[124],"our":[125],"findings":[126],"challenge":[127],"growing":[129],"assumption":[130],"research,":[133],"namely":[134],"progress":[136,141],"generative":[138],"realism":[139],"implies":[140],"realism.":[144],"thus":[146],"highlight":[147],"an":[148],"urgent":[149],"need":[150],"rethink":[152],"capabilities":[154],"modern":[156],"reliable":[160]},"counts_by_year":[],"updated_date":"2026-07-28T07:46:37.118299","created_date":"2026-02-26T00:00:00"}