{"id":"https://openalex.org/W4413157714","doi":"https://doi.org/10.1109/cvpr52734.2025.00751","title":"Personalized Preference Fine-tuning of Diffusion Models","display_name":"Personalized Preference Fine-tuning of Diffusion Models","publication_year":2025,"publication_date":"2025-06-10","ids":{"openalex":"https://openalex.org/W4413157714","doi":"https://doi.org/10.1109/cvpr52734.2025.00751"},"language":"en","primary_location":{"id":"doi:10.1109/cvpr52734.2025.00751","is_oa":false,"landing_page_url":"https://doi.org/10.1109/cvpr52734.2025.00751","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2025 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)","raw_type":"proceedings-article"},"type":"article","indexed_in":["crossref"],"open_access":{"is_oa":false,"oa_status":"closed","oa_url":null,"any_repository_has_fulltext":false},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5043265692","display_name":"Meihua Dang","orcid":"https://orcid.org/0000-0001-8241-1943"},"institutions":[{"id":"https://openalex.org/I97018004","display_name":"Stanford University","ror":"https://ror.org/00f54p054","country_code":"US","type":"education","lineage":["https://openalex.org/I97018004"]}],"countries":["US"],"is_corresponding":true,"raw_author_name":"Meihua Dang","raw_affiliation_strings":["Stanford University"],"affiliations":[{"raw_affiliation_string":"Stanford University","institution_ids":["https://openalex.org/I97018004"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5088818925","display_name":"Anikait Singh","orcid":null},"institutions":[{"id":"https://openalex.org/I97018004","display_name":"Stanford University","ror":"https://ror.org/00f54p054","country_code":"US","type":"education","lineage":["https://openalex.org/I97018004"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Anikait Singh","raw_affiliation_strings":["Stanford University"],"affiliations":[{"raw_affiliation_string":"Stanford University","institution_ids":["https://openalex.org/I97018004"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5074297682","display_name":"Linqi Zhou","orcid":"https://orcid.org/0009-0003-8981-2661"},"institutions":[{"id":"https://openalex.org/I97018004","display_name":"Stanford University","ror":"https://ror.org/00f54p054","country_code":"US","type":"education","lineage":["https://openalex.org/I97018004"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Linqi Zhou","raw_affiliation_strings":["Stanford University"],"affiliations":[{"raw_affiliation_string":"Stanford University","institution_ids":["https://openalex.org/I97018004"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5113804858","display_name":"Stefano Ermon","orcid":null},"institutions":[{"id":"https://openalex.org/I97018004","display_name":"Stanford University","ror":"https://ror.org/00f54p054","country_code":"US","type":"education","lineage":["https://openalex.org/I97018004"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Stefano Ermon","raw_affiliation_strings":["Stanford University"],"affiliations":[{"raw_affiliation_string":"Stanford University","institution_ids":["https://openalex.org/I97018004"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5008049062","display_name":"Jiaming Song","orcid":"https://orcid.org/0000-0002-1931-1699"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Jiaming Song","raw_affiliation_strings":["Luma AI"],"affiliations":[{"raw_affiliation_string":"Luma AI","institution_ids":[]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":5,"corresponding_author_ids":["https://openalex.org/A5043265692"],"corresponding_institution_ids":["https://openalex.org/I97018004"],"apc_list":null,"apc_paid":null,"fwci":2.4307,"has_fulltext":false,"cited_by_count":2,"citation_normalized_percentile":{"value":0.89359666,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":{"min":91,"max":99},"biblio":{"volume":null,"issue":null,"first_page":"8020","last_page":"8030"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T12114","display_name":"Sensory Analysis and Statistical Methods","score":0.3504999876022339,"subfield":{"id":"https://openalex.org/subfields/1106","display_name":"Food Science"},"field":{"id":"https://openalex.org/fields/11","display_name":"Agricultural and Biological Sciences"},"domain":{"id":"https://openalex.org/domains/1","display_name":"Life Sciences"}},"topics":[{"id":"https://openalex.org/T12114","display_name":"Sensory Analysis and Statistical Methods","score":0.3504999876022339,"subfield":{"id":"https://openalex.org/subfields/1106","display_name":"Food Science"},"field":{"id":"https://openalex.org/fields/11","display_name":"Agricultural and Biological Sciences"},"domain":{"id":"https://openalex.org/domains/1","display_name":"Life Sciences"}},{"id":"https://openalex.org/T11106","display_name":"Data Management and Algorithms","score":0.34619998931884766,"subfield":{"id":"https://openalex.org/subfields/1711","display_name":"Signal Processing"},"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":0.6312225461006165},{"id":"https://openalex.org/keywords/preference","display_name":"Preference","score":0.5657390356063843},{"id":"https://openalex.org/keywords/diffusion","display_name":"Diffusion","score":0.5450729131698608},{"id":"https://openalex.org/keywords/physics","display_name":"Physics","score":0.11650577187538147},{"id":"https://openalex.org/keywords/mathematics","display_name":"Mathematics","score":0.08908075094223022},{"id":"https://openalex.org/keywords/statistics","display_name":"Statistics","score":0.07087203860282898}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6312225461006165},{"id":"https://openalex.org/C2781249084","wikidata":"https://www.wikidata.org/wiki/Q908656","display_name":"Preference","level":2,"score":0.5657390356063843},{"id":"https://openalex.org/C69357855","wikidata":"https://www.wikidata.org/wiki/Q163214","display_name":"Diffusion","level":2,"score":0.5450729131698608},{"id":"https://openalex.org/C121332964","wikidata":"https://www.wikidata.org/wiki/Q413","display_name":"Physics","level":0,"score":0.11650577187538147},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.08908075094223022},{"id":"https://openalex.org/C105795698","wikidata":"https://www.wikidata.org/wiki/Q12483","display_name":"Statistics","level":1,"score":0.07087203860282898},{"id":"https://openalex.org/C97355855","wikidata":"https://www.wikidata.org/wiki/Q11473","display_name":"Thermodynamics","level":1,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/cvpr52734.2025.00751","is_oa":false,"landing_page_url":"https://doi.org/10.1109/cvpr52734.2025.00751","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2025 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)","raw_type":"proceedings-article"}],"best_oa_location":null,"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":11,"referenced_works":["https://openalex.org/W2763110165","https://openalex.org/W4312933868","https://openalex.org/W4390873054","https://openalex.org/W4390874002","https://openalex.org/W4393148714","https://openalex.org/W4402670365","https://openalex.org/W4402753770","https://openalex.org/W4402753790","https://openalex.org/W4404545718","https://openalex.org/W4404783047","https://openalex.org/W4408075062"],"related_works":["https://openalex.org/W4391375266","https://openalex.org/W2899084033","https://openalex.org/W2748952813","https://openalex.org/W2390279801","https://openalex.org/W4391913857","https://openalex.org/W2358668433","https://openalex.org/W4396701345","https://openalex.org/W2376932109","https://openalex.org/W2001405890","https://openalex.org/W4396696052"],"abstract_inverted_index":{"RLHF":[0],"techniques":[1],"like":[2],"DPO":[3,134],"can":[4,159],"significantly":[5],"improve":[6],"the":[7,30,43,72,115,127,133,141],"generation":[8,25],"quality":[9],"of":[10,32,40,45,75,78,107,143,187],"text-to-image":[11,128],"diffusion":[12,61,69,118],"models.":[13,47],"However,":[14],"these":[15,46],"methods":[16],"optimize":[17],"for":[18,138,154],"a":[19,55,68,76,81,94,104,177],"single":[20],"reward":[21,156],"that":[22,59,149,194],"aligns":[23,60],"model":[24,70,96],"with":[26,63,132,140,169],"population-level":[27],"preferences,":[28],"neglecting":[29],"nuances":[31],"individual":[33,73],"users\u2019":[34],"beliefs":[35],"or":[36],"values.":[37],"This":[38],"lack":[39],"personalization":[41],"limits":[42],"efficacy":[44],"To":[48],"bridge":[49],"this":[50],"gap,":[51],"we":[52],"introduce":[53],"PPD,":[54,67],"multi-reward":[56],"optimization":[57],"objective":[58],"models":[62,119,129],"personalized":[64],"preferences.":[65,200],"With":[66],"learns":[71],"preferences":[74,142],"population":[77],"users":[79],"in":[80],"few-shot":[82],"way,":[83],"enabling":[84],"generalization":[85],"to":[86,98],"unseen":[87],"users.":[88,145],"Specifically,":[89],"our":[90,150,180],"approach":[91,181],"(1)":[92],"leverages":[93],"vision-language":[95],"(VLM)":[97],"extract":[99],"personal":[100],"preference":[101,109,174],"embeddings":[102,116],"from":[103,176],"small":[105],"set":[106],"pairwise":[108],"examples,":[110],"and":[111,158],"then":[112],"(2)":[113],"incorporates":[114],"into":[117],"through":[120],"cross":[121],"attention.":[122],"Conditioning":[123],"on":[124],"user":[125,167,199],"embeddings,":[126],"are":[130],"fine-tuned":[131],"objective,":[135],"simultaneously":[136],"optimizing":[137],"alignment":[139],"multiple":[144,155],"Empirical":[146],"results":[147],"demonstrate":[148],"method":[151],"effectively":[152],"optimizes":[153],"functions":[157],"interpolate":[160],"between":[161],"them":[162],"during":[163],"inference.":[164],"In":[165],"real-world":[166],"scenarios,":[168],"as":[170,172],"few":[171],"four":[173],"examples":[175],"new":[178],"user,":[179],"achieves":[182],"an":[183],"average":[184],"win":[185],"rate":[186],"76%":[188],"over":[189],"Stable":[190],"Cascade,":[191],"generating":[192],"images":[193],"more":[195],"accurately":[196],"reflect":[197],"specific":[198]},"counts_by_year":[{"year":2026,"cited_by_count":1},{"year":2025,"cited_by_count":1}],"updated_date":"2025-12-21T01:58:51.020947","created_date":"2025-10-10T00:00:00"}