{"id":"https://openalex.org/W7080415457","doi":"https://doi.org/10.48550/arxiv.2509.06942","title":"Directly Aligning the Full Diffusion Trajectory with Fine-Grained Human Preference","display_name":"Directly Aligning the Full Diffusion Trajectory with Fine-Grained Human Preference","publication_year":2025,"publication_date":"2025-09-08","ids":{"openalex":"https://openalex.org/W7080415457","doi":"https://doi.org/10.48550/arxiv.2509.06942"},"language":"en","primary_location":{"id":"pmh:oai:arXiv.org:2509.06942","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2509.06942","pdf_url":"https://arxiv.org/pdf/2509.06942","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":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"volume":null,"type":"preprint","is_traditional":true,"indexed_in":["arxiv","datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://arxiv.org/pdf/2509.06942","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":null,"display_name":"Shen, Xiangwei","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I177725633","display_name":"Chinese University of Hong Kong","ror":"https://ror.org/00t33hh48","country_code":"HK","type":"education","lineage":["https://openalex.org/I177725633"]},{"id":"https://openalex.org/I4210116924","display_name":"Chinese University of Hong Kong, Shenzhen","ror":"https://ror.org/02d5ks197","country_code":"CN","type":"education","lineage":["https://openalex.org/I177725633","https://openalex.org/I180726961","https://openalex.org/I4210116924"]}],"countries":["CN","HK"],"is_corresponding":false,"raw_author_name":"Shen, Xiangwei","raw_affiliation_strings":["School of Science and Engineering , The Chinese University of Hong Kong , Shenzhen"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"School of Science and Engineering , The Chinese University of Hong Kong , Shenzhen","institution_ids":["https://openalex.org/I177725633","https://openalex.org/I4210116924"]}]},{"author_position":"middle","author":{"id":null,"display_name":"Li, Zhimin","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Li, Zhimin","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":null,"display_name":"Yang, Zhantao","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yang, Zhantao","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":null,"display_name":"Zhang, Shiyi","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I4405263052","display_name":"Tsinghua Shenzhen International Graduate School","ror":"https://ror.org/01vceef04","country_code":"CN","type":"education","lineage":["https://openalex.org/I4405263052","https://openalex.org/I99065089"]},{"id":"https://openalex.org/I99065089","display_name":"Tsinghua University","ror":"https://ror.org/03cve4549","country_code":"CN","type":"education","lineage":["https://openalex.org/I99065089"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Zhang, Shiyi","raw_affiliation_strings":["Shenzhen International Graduate School , Tsinghua University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Shenzhen International Graduate School , Tsinghua University","institution_ids":["https://openalex.org/I4405263052","https://openalex.org/I99065089"]}]},{"author_position":"middle","author":{"id":null,"display_name":"Zhang, Yingfang","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhang, Yingfang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":null,"display_name":"Li, Donghao","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Li, Donghao","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":null,"display_name":"Wang, Chunyu","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wang, Chunyu","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":null,"display_name":"Lu, Qinglin","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Lu, Qinglin","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":null,"display_name":"Tang, Yansong","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I4405263052","display_name":"Tsinghua Shenzhen International Graduate School","ror":"https://ror.org/01vceef04","country_code":"CN","type":"education","lineage":["https://openalex.org/I4405263052","https://openalex.org/I99065089"]},{"id":"https://openalex.org/I99065089","display_name":"Tsinghua University","ror":"https://ror.org/03cve4549","country_code":"CN","type":"education","lineage":["https://openalex.org/I99065089"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Tang, Yansong","raw_affiliation_strings":["Shenzhen International Graduate School , Tsinghua University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Shenzhen International Graduate School , Tsinghua University","institution_ids":["https://openalex.org/I4405263052","https://openalex.org/I99065089"]}]}],"institutions":[],"countries_distinct_count":2,"institutions_distinct_count":4,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":true,"cited_by_count":26,"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":true,"primary_topic":{"id":"https://openalex.org/T10775","display_name":"Generative Adversarial Networks and Image Synthesis","score":1.0,"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"}},"topics":[{"id":"https://openalex.org/T10775","display_name":"Generative Adversarial Networks and Image Synthesis","score":1.0,"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"}},{"id":"https://openalex.org/T11689","display_name":"Adversarial Robustness in Machine Learning","score":0.0,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"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/T10481","display_name":"Computer Graphics and Visualization Techniques","score":0.0,"subfield":{"id":"https://openalex.org/subfields/1704","display_name":"Computer Graphics and Computer-Aided Design"},"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/human-preferences","display_name":"human preferences","score":0.5809999704360962},{"id":"https://openalex.org/keywords/prompt-augmentation","display_name":"prompt augmentation","score":0.6949999928474426},{"id":"https://openalex.org/keywords/flux-model","display_name":"flux model","score":0.8050000071525574},{"id":"https://openalex.org/keywords/photorealism","display_name":"photorealism","score":0.7099999785423279},{"id":"https://openalex.org/keywords/aesthetic-quality","display_name":"aesthetic quality","score":0.6729999780654907},{"id":"https://openalex.org/keywords/fine-tuning","display_name":"fine-tuning","score":0.3799999952316284}],"concepts":[{"id":"https://openalex.org/C99498987","wikidata":"https://www.wikidata.org/wiki/Q2210247","display_name":"Noise (video)","level":3,"score":0.661300003528595},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6308000087738037},{"id":"https://openalex.org/C163294075","wikidata":"https://www.wikidata.org/wiki/Q581861","display_name":"Noise reduction","level":2,"score":0.6219000220298767},{"id":"https://openalex.org/C45374587","wikidata":"https://www.wikidata.org/wiki/Q12525525","display_name":"Computation","level":2,"score":0.5889999866485596},{"id":"https://openalex.org/C13662910","wikidata":"https://www.wikidata.org/wiki/Q193139","display_name":"Trajectory","level":2,"score":0.5507000088691711},{"id":"https://openalex.org/C2781249084","wikidata":"https://www.wikidata.org/wiki/Q908656","display_name":"Preference","level":2,"score":0.5115000009536743},{"id":"https://openalex.org/C202615002","wikidata":"https://www.wikidata.org/wiki/Q783507","display_name":"Differentiable function","level":2,"score":0.478300005197525},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.47029998898506165},{"id":"https://openalex.org/C126255220","wikidata":"https://www.wikidata.org/wiki/Q141495","display_name":"Mathematical optimization","level":1,"score":0.45980000495910645},{"id":"https://openalex.org/C139807058","wikidata":"https://www.wikidata.org/wiki/Q352374","display_name":"Adaptation (eye)","level":2,"score":0.4438000023365021},{"id":"https://openalex.org/C69357855","wikidata":"https://www.wikidata.org/wiki/Q163214","display_name":"Diffusion","level":2,"score":0.42260000109672546},{"id":"https://openalex.org/C2779530757","wikidata":"https://www.wikidata.org/wiki/Q1207505","display_name":"Quality (philosophy)","level":2,"score":0.4049000144004822},{"id":"https://openalex.org/C2777851325","wikidata":"https://www.wikidata.org/wiki/Q7094102","display_name":"Online model","level":2,"score":0.39480000734329224},{"id":"https://openalex.org/C111335779","wikidata":"https://www.wikidata.org/wiki/Q3454686","display_name":"Reduction (mathematics)","level":2,"score":0.35679998993873596},{"id":"https://openalex.org/C137836250","wikidata":"https://www.wikidata.org/wiki/Q984063","display_name":"Optimization problem","level":2,"score":0.3294999897480011},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.3052000105381012},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.30160000920295715},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.2913999855518341},{"id":"https://openalex.org/C182306322","wikidata":"https://www.wikidata.org/wiki/Q1779371","display_name":"Order (exchange)","level":2,"score":0.2888000011444092},{"id":"https://openalex.org/C29265498","wikidata":"https://www.wikidata.org/wiki/Q7047719","display_name":"Noise measurement","level":3,"score":0.2761000096797943},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.2669999897480011},{"id":"https://openalex.org/C31972630","wikidata":"https://www.wikidata.org/wiki/Q844240","display_name":"Computer vision","level":1,"score":0.25429999828338623},{"id":"https://openalex.org/C47446073","wikidata":"https://www.wikidata.org/wiki/Q5165890","display_name":"Control theory (sociology)","level":3,"score":0.2529999911785126}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:arXiv.org:2509.06942","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2509.06942","pdf_url":"https://arxiv.org/pdf/2509.06942","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":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},{"id":"doi:10.48550/arxiv.2509.06942","doi":"https://doi.org/10.48550/arxiv.2509.06942","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2509.06942","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":null,"license_id":null,"version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"pmh:oai:arXiv.org:2509.06942","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2509.06942","pdf_url":"https://arxiv.org/pdf/2509.06942","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":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"sustainable_development_goals":[],"study_designs":[],"awards":[],"funders":[],"has_content":{"pdf":true,"grobid_xml":true},"content_urls":{"pdf":"https://content.openalex.org/works/W7080415457.pdf","grobid_xml":"https://content.openalex.org/works/W7080415457.grobid-xml"},"referenced_works_count":22,"referenced_works":["https://openalex.org/W3092442149","https://openalex.org/W3110257065","https://openalex.org/W4226251505","https://openalex.org/W4288099666","https://openalex.org/W4297676498","https://openalex.org/W4303647933","https://openalex.org/W4318903518","https://openalex.org/W4321855132","https://openalex.org/W4327673663","https://openalex.org/W4378464471","https://openalex.org/W4385570204","https://openalex.org/W4387294554","https://openalex.org/W4396945138","https://openalex.org/W4400612128","https://openalex.org/W4403662179","https://openalex.org/W4407128718","https://openalex.org/W4413145151","https://openalex.org/W4414449548","https://openalex.org/W4416070669","https://openalex.org/W4417517408","https://openalex.org/W6929198645","https://openalex.org/W7160158350"],"related_works":["https://openalex.org/W4407184534","https://openalex.org/W4403347013","https://openalex.org/W7138224807","https://openalex.org/W4402753790","https://openalex.org/W7117703275","https://openalex.org/W4400065145","https://openalex.org/W6966766647","https://openalex.org/W7162408510","https://openalex.org/W4416189736","https://openalex.org/W4412598373"],"abstract_inverted_index":{"Recent":[0],"studies":[1],"have":[2],"demonstrated":[3],"the":[4,74,101,153,161],"effectiveness":[5],"of":[6,55,76,141],"directly":[7],"aligning":[8],"diffusion":[9,46,104],"models":[10,57],"with":[11,29,164],"human":[12],"preferences":[13],"using":[14],"differentiable":[15],"reward.":[16],"However,":[17],"they":[18,24,49],"exhibit":[19],"two":[20],"primary":[21],"challenges:":[22],"(1)":[23],"rely":[25],"on":[26,155],"multistep":[27,77],"denoising":[28,166],"gradient":[30],"computation":[31],"for":[32],"reward":[33,56,157,169],"scoring,":[34],"which":[35,113,129],"is":[36],"computationally":[37],"expensive,":[38],"thus":[39],"restricting":[40],"optimization":[41],"to":[42,60,89,145],"only":[43],"a":[44,82,86],"few":[45],"steps;":[47],"(2)":[48],"often":[50],"need":[51],"continuous":[52],"offline":[53,156],"adaptation":[54],"in":[58,117,128,143],"order":[59],"achieve":[61],"desired":[62],"aesthetic":[63,177],"quality,":[64],"such":[65],"as":[66,133],"photorealism":[67],"or":[68],"precise":[69],"lighting":[70],"effects.":[71],"To":[72],"address":[73],"limitation":[75],"denoising,":[78],"we":[79,121,171],"propose":[80],"Direct-Align,":[81],"method":[83],"that":[84,103],"predefines":[85],"noise":[87,109],"prior":[88],"effectively":[90,114],"recover":[91],"original":[92],"images":[93],"from":[94],"any":[95],"time":[96],"steps":[97],"via":[98],"interpolation,":[99],"leveraging":[100],"equation":[102],"states":[105],"are":[106,131],"interpolations":[107],"between":[108],"and":[110,147,167,176],"target":[111],"images,":[112],"avoids":[115],"over-optimization":[116],"late":[118],"timesteps.":[119],"Furthermore,":[120],"introduce":[122],"Semantic":[123],"Relative":[124],"Preference":[125],"Optimization":[126],"(SRPO),":[127],"rewards":[130,142],"formulated":[132],"text-conditioned":[134],"signals.":[135],"This":[136],"approach":[137],"enables":[138],"online":[139,168],"adjustment":[140],"response":[144],"positive":[146],"negative":[148],"prompt":[149],"augmentation,":[150],"thereby":[151],"reducing":[152],"reliance":[154],"fine-tuning.":[158],"By":[159],"fine-tuning":[160],"FLUX":[162],"model":[163],"optimized":[165],"adjustment,":[170],"improve":[172],"its":[173],"human-evaluated":[174],"realism":[175],"quality":[178],"by":[179],"over":[180],"3x.":[181]},"counts_by_year":[{"year":2026,"cited_by_count":21},{"year":2025,"cited_by_count":5}],"updated_date":"2026-11-07T11:25:07.637180","created_date":"2025-10-10T00:00:00"}