{"id":"https://openalex.org/W7166321509","doi":"https://doi.org/10.5281/zenodo.20952260","title":"What Transfers and What Collapses: A Cross-Generator Study of AI-Generated Image Detection with Corrected Evaluation","display_name":"What Transfers and What Collapses: A Cross-Generator Study of AI-Generated Image Detection with Corrected Evaluation","publication_year":2026,"publication_date":"2026-06-27","ids":{"openalex":"https://openalex.org/W7166321509","doi":"https://doi.org/10.5281/zenodo.20952260"},"language":"en","primary_location":{"id":"doi:10.5281/zenodo.20952260","doi":"https://doi.org/10.5281/zenodo.20952260","is_oa":true,"landing_page_url":"https://doi.org/10.5281/zenodo.20952260","pdf_url":null,"source":{"id":"https://openalex.org/S4306400562","display_name":"Zenodo (CERN European Organization for Nuclear Research)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I67311998","host_organization_name":"European Organization for Nuclear Research","host_organization_lineage":["https://openalex.org/I67311998"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"},"volume":null,"type":"preprint","indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://doi.org/10.5281/zenodo.20952260","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5138192180","display_name":"Mohamed Alaya","orcid":"https://orcid.org/0009-0002-1668-1923","observed_orcids":["https://orcid.org/0009-0002-1668-1923"]},"institutions":[],"countries":[],"is_corresponding":true,"raw_author_name":"Alaya, Mohamed","raw_affiliation_strings":["ISIMG, Tunisia"],"raw_orcid":"https://orcid.org/0009-0002-1668-1923","affiliations":[{"raw_affiliation_string":"ISIMG, Tunisia","institution_ids":[]}]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":0,"corresponding_author_ids":["https://openalex.org/A5138192180"],"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/T12357","display_name":"Digital Media Forensic Detection","score":0.9990000128746033,"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/T12357","display_name":"Digital Media Forensic Detection","score":0.9990000128746033,"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/T10775","display_name":"Generative Adversarial Networks and Image Synthesis","score":0.0005000000237487257,"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/T10388","display_name":"Advanced Steganography and Watermarking Techniques","score":9.999999747378752e-05,"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/ai-generated-image-detection","display_name":"AI-generated image detection","score":0.9800000190734863},{"id":"https://openalex.org/keywords/stable-diffusion","display_name":"Stable Diffusion","score":0.5849999785423279},{"id":"https://openalex.org/keywords/diffusion-models","display_name":"diffusion models","score":0.6430000066757202},{"id":"https://openalex.org/keywords/gan","display_name":"GaN","score":0.8019999861717224},{"id":"https://openalex.org/keywords/clip-features","display_name":"CLIP features","score":0.5740000009536743},{"id":"https://openalex.org/keywords/frozen-foundation-models","display_name":"frozen foundation models","score":0.5460000038146973},{"id":"https://openalex.org/keywords/adversarial-audit","display_name":"adversarial audit","score":0.4779999852180481},{"id":"https://openalex.org/keywords/bootstrap-confidence-intervals","display_name":"bootstrap confidence intervals","score":0.5960000157356262},{"id":"https://openalex.org/keywords/dall-e","display_name":"DALL-E","score":0.4830000102519989}],"concepts":[{"id":"https://openalex.org/C94915269","wikidata":"https://www.wikidata.org/wiki/Q1834857","display_name":"Detector","level":2,"score":0.7476000189781189},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.6021999716758728},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6000000238418579},{"id":"https://openalex.org/C31972630","wikidata":"https://www.wikidata.org/wiki/Q844240","display_name":"Computer vision","level":1,"score":0.5063999891281128},{"id":"https://openalex.org/C185798385","wikidata":"https://www.wikidata.org/wiki/Q1161707","display_name":"Benchmark (surveying)","level":2,"score":0.49219998717308044},{"id":"https://openalex.org/C160633673","wikidata":"https://www.wikidata.org/wiki/Q355198","display_name":"Pixel","level":2,"score":0.4180999994277954},{"id":"https://openalex.org/C153180895","wikidata":"https://www.wikidata.org/wiki/Q7148389","display_name":"Pattern recognition (psychology)","level":2,"score":0.4113999903202057},{"id":"https://openalex.org/C2776760102","wikidata":"https://www.wikidata.org/wiki/Q5139990","display_name":"Code (set theory)","level":3,"score":0.4034999907016754},{"id":"https://openalex.org/C115961682","wikidata":"https://www.wikidata.org/wiki/Q860623","display_name":"Image (mathematics)","level":2,"score":0.3668999969959259},{"id":"https://openalex.org/C2780385302","wikidata":"https://www.wikidata.org/wiki/Q367158","display_name":"Protocol (science)","level":3,"score":0.3398999869823456},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.2980000078678131},{"id":"https://openalex.org/C2776151529","wikidata":"https://www.wikidata.org/wiki/Q3045304","display_name":"Object detection","level":3,"score":0.26269999146461487},{"id":"https://openalex.org/C2780586882","wikidata":"https://www.wikidata.org/wiki/Q7520643","display_name":"Simple (philosophy)","level":2,"score":0.2614000141620636},{"id":"https://openalex.org/C2776359362","wikidata":"https://www.wikidata.org/wiki/Q2145286","display_name":"Representation (politics)","level":3,"score":0.2612000107765198},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.2529999911785126}],"mesh":[],"locations_count":4,"locations":[{"id":"doi:10.5281/zenodo.20952260","doi":"https://doi.org/10.5281/zenodo.20952260","is_oa":true,"landing_page_url":"https://doi.org/10.5281/zenodo.20952260","pdf_url":null,"source":{"id":"https://openalex.org/S4306400562","display_name":"Zenodo (CERN European Organization for Nuclear Research)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I67311998","host_organization_name":"European Organization for Nuclear Research","host_organization_lineage":["https://openalex.org/I67311998"],"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"},{"id":"doi:10.5281/zenodo.20952261","doi":"https://doi.org/10.5281/zenodo.20952261","is_oa":true,"landing_page_url":"https://doi.org/10.5281/zenodo.20952261","pdf_url":null,"source":{"id":"https://openalex.org/S4306400562","display_name":"Zenodo (CERN European Organization for Nuclear Research)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I67311998","host_organization_name":"European Organization for Nuclear Research","host_organization_lineage":["https://openalex.org/I67311998"],"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"},{"id":"doi:10.5281/zenodo.22692235","doi":"https://doi.org/10.5281/zenodo.22692235","is_oa":true,"landing_page_url":"https://doi.org/10.5281/zenodo.22692235","pdf_url":null,"source":{"id":"https://openalex.org/S4306400562","display_name":"Zenodo (CERN European Organization for Nuclear Research)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I67311998","host_organization_name":"European Organization for Nuclear Research","host_organization_lineage":["https://openalex.org/I67311998"],"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"},{"id":"doi:10.5281/zenodo.22692735","doi":"https://doi.org/10.5281/zenodo.22692735","is_oa":true,"landing_page_url":"https://doi.org/10.5281/zenodo.22692735","pdf_url":null,"source":{"id":"https://openalex.org/S4306400562","display_name":"Zenodo (CERN European Organization for Nuclear Research)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I67311998","host_organization_name":"European Organization for Nuclear Research","host_organization_lineage":["https://openalex.org/I67311998"],"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":"doi:10.5281/zenodo.20952260","doi":"https://doi.org/10.5281/zenodo.20952260","is_oa":true,"landing_page_url":"https://doi.org/10.5281/zenodo.20952260","pdf_url":null,"source":{"id":"https://openalex.org/S4306400562","display_name":"Zenodo (CERN European Organization for Nuclear Research)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I67311998","host_organization_name":"European Organization for Nuclear Research","host_organization_lineage":["https://openalex.org/I67311998"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"},"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":[],"abstract_inverted_index":{"A":[0,122],"cross-generator":[1,46],"study":[2],"of":[3,11],"AI-generated":[4],"image":[5],"detection":[6],"on":[7,58,68,129,134,181],"a":[8,29,48,62,73,93,161],"14-generator":[9],"benchmark":[10],"modern":[12],"systems":[13],"(Stable":[14],"Diffusion":[15],"1.3/1.4/2/XL,":[16],"SD3,":[17],"FLUX.1-dev/schnell,":[18],"DALL-E":[19],"2/3,":[20],"Midjourney":[21],"v5,":[22],"Imagen":[23],"3,":[24],"GLIDE,":[25],"Adobe":[26],"Firefly),":[27],"under":[28],"leak-free":[30],"leave-generators-out":[31],"protocol":[32],"with":[33,177],"corrected":[34],"metrics":[35],"and":[36,61,92,116,172,195,200],"bootstrap":[37],"confidence":[38],"intervals.":[39],"Findings.":[40],"In-distribution":[41],"accuracy":[42],"does":[43],"not":[44,149],"predict":[45],"accuracy:":[47],"2-D":[49],"spectral":[50],"detector":[51,76],"scores":[52],"0.795":[53],"in-distribution":[54],"but":[55,154],"0.523":[56],"(chance)":[57],"unseen":[59,71,135],"generators,":[60],"fine-tuned":[63],"CNN":[64,113],"drops":[65],"to":[66,114,120],"chance":[67,80],"its":[69],"worst":[70],"generator;":[72],"hand-crafted":[74],"physics":[75],"goes":[77],"confidently":[78],"below":[79],"(0.258,":[81],"fingerprint":[82],"inversion).":[83],"Only":[84],"frozen":[85],"foundation":[86],"features":[87],"generalize":[88],"(frozen-CLIP":[89],"probe,":[90,153],"0.917),":[91],"one-class":[94,144],"real-manifold":[95],"model":[96],"structurally":[97],"cannot":[98],"sign-invert":[99],"(worst-case":[100],"floor":[101],"0.611).":[102],"The":[103],"collapse":[104],"is":[105],"governed":[106],"by":[107,160],"training-generator":[108],"diversity":[109],"(leave-one-out":[110],"recovers":[111],"the":[112,117,151,178,182],"0.876":[115],"CLIP":[118],"probe":[119,124],"0.960).":[121],"cross-architecture":[123],"shows":[125],"detectors":[126],"do":[127,148],"better":[128],"an":[130,188],"older":[131],"GAN":[132],"than":[133],"diffusion.":[136],"Improvement.":[137],"Feature-space":[138],"extensions":[139],"(DINOv2":[140],"ensembling,":[141],"generator-direction":[142],"removal,":[143],"fusion,":[145],"reconstruction":[146],"fusion)":[147],"beat":[150],"frozen-CLIP":[152],"simple":[155],"5-crop":[156],"test-time":[157],"aggregation":[158],"does,":[159],"paired-bootstrap-significant":[162],"margin:":[163],"0.914":[164],"->":[165,174],"0.949":[166],"(three-seen,":[167],"95%":[168],"CI":[169],"[0.028,":[170],"0.042])":[171],"0.961":[173],"0.972":[175],"(leave-one-out),":[176],"largest":[179],"gains":[180],"hardest":[183],"generators.":[184],"We":[185],"also":[186],"document":[187],"AUC-orientation":[189],"evaluation":[190],"bug":[191],"that":[192],"returns":[193],"1-AUC":[194],"silently":[196],"inverts":[197],"results.":[198],"Code":[199],"benchmark:":[201],"https://github.com/theFinex/cross-generator-ai-detection":[202]},"counts_by_year":[],"updated_date":"2026-10-08T09:06:37.294493","created_date":"2026-06-28T00:00:00"}