{"id":"https://openalex.org/W7158681843","doi":"https://doi.org/10.48550/arxiv.2604.26503","title":"Delta Score Matters! Spatial Adaptive Multi Guidance in Diffusion Models","display_name":"Delta Score Matters! Spatial Adaptive Multi Guidance in Diffusion Models","publication_year":2026,"publication_date":"2026-04-29","ids":{"openalex":"https://openalex.org/W7158681843","doi":"https://doi.org/10.48550/arxiv.2604.26503"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2604.26503","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2604.26503","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"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":false,"raw_source_name":null,"raw_type":"Preprint"},"type":"preprint","indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://doi.org/10.48550/arxiv.2604.26503","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5134888372","display_name":"Haosen Li","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Li, Haosen","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5134928562","display_name":"Wenshuo Chen","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Chen, Wenshuo","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5134901943","display_name":"Lei Wang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wang, Lei","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5134895320","display_name":"Shaofeng Liang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Liang, Shaofeng","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5134918541","display_name":"Bowen Tian","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Tian, Bowen","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5134922584","display_name":"Soning Lai","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Lai, Soning","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5134877036","display_name":"Yutao Yue","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yue, Yutao","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":0.8360000252723694,"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":0.8360000252723694,"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/T11307","display_name":"Domain Adaptation and Few-Shot Learning","score":0.03359999880194664,"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/T11714","display_name":"Multimodal Machine Learning Applications","score":0.03139999881386757,"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/smoothness","display_name":"Smoothness","score":0.5695000290870667},{"id":"https://openalex.org/keywords/trajectory","display_name":"Trajectory","score":0.5332000255584717},{"id":"https://openalex.org/keywords/scale","display_name":"Scale (ratio)","score":0.5192000269889832},{"id":"https://openalex.org/keywords/boundary","display_name":"Boundary (topology)","score":0.5027999877929688},{"id":"https://openalex.org/keywords/sampling","display_name":"Sampling (signal processing)","score":0.4171000123023987},{"id":"https://openalex.org/keywords/adaptive-sampling","display_name":"Adaptive sampling","score":0.4081999957561493},{"id":"https://openalex.org/keywords/manifold","display_name":"Manifold (fluid mechanics)","score":0.3628999888896942},{"id":"https://openalex.org/keywords/measure","display_name":"Measure (data warehouse)","score":0.36239999532699585},{"id":"https://openalex.org/keywords/diffusion","display_name":"Diffusion","score":0.35659998655319214}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6513000130653381},{"id":"https://openalex.org/C102634674","wikidata":"https://www.wikidata.org/wiki/Q868473","display_name":"Smoothness","level":2,"score":0.5695000290870667},{"id":"https://openalex.org/C13662910","wikidata":"https://www.wikidata.org/wiki/Q193139","display_name":"Trajectory","level":2,"score":0.5332000255584717},{"id":"https://openalex.org/C2778755073","wikidata":"https://www.wikidata.org/wiki/Q10858537","display_name":"Scale (ratio)","level":2,"score":0.5192000269889832},{"id":"https://openalex.org/C62354387","wikidata":"https://www.wikidata.org/wiki/Q875399","display_name":"Boundary (topology)","level":2,"score":0.5027999877929688},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.5020999908447266},{"id":"https://openalex.org/C140779682","wikidata":"https://www.wikidata.org/wiki/Q210868","display_name":"Sampling (signal processing)","level":3,"score":0.4171000123023987},{"id":"https://openalex.org/C2781395549","wikidata":"https://www.wikidata.org/wiki/Q4680762","display_name":"Adaptive sampling","level":3,"score":0.4081999957561493},{"id":"https://openalex.org/C529865628","wikidata":"https://www.wikidata.org/wiki/Q1790740","display_name":"Manifold (fluid mechanics)","level":2,"score":0.3628999888896942},{"id":"https://openalex.org/C2780009758","wikidata":"https://www.wikidata.org/wiki/Q6804172","display_name":"Measure (data warehouse)","level":2,"score":0.36239999532699585},{"id":"https://openalex.org/C69357855","wikidata":"https://www.wikidata.org/wiki/Q163214","display_name":"Diffusion","level":2,"score":0.35659998655319214},{"id":"https://openalex.org/C93226319","wikidata":"https://www.wikidata.org/wiki/Q193137","display_name":"Differential (mechanical device)","level":2,"score":0.35530000925064087},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.3529999852180481},{"id":"https://openalex.org/C77553402","wikidata":"https://www.wikidata.org/wiki/Q13222579","display_name":"Upper and lower bounds","level":2,"score":0.3285999894142151},{"id":"https://openalex.org/C2780801425","wikidata":"https://www.wikidata.org/wiki/Q5164392","display_name":"Construct (python library)","level":2,"score":0.2955000102519989},{"id":"https://openalex.org/C66882249","wikidata":"https://www.wikidata.org/wiki/Q169336","display_name":"Homogeneous","level":2,"score":0.2897999882698059},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.28850001096725464},{"id":"https://openalex.org/C89600930","wikidata":"https://www.wikidata.org/wiki/Q1423946","display_name":"Segmentation","level":2,"score":0.28839999437332153},{"id":"https://openalex.org/C2987819851","wikidata":"https://www.wikidata.org/wiki/Q191839","display_name":"Aerial imagery","level":2,"score":0.2874999940395355},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.28630000352859497},{"id":"https://openalex.org/C31972630","wikidata":"https://www.wikidata.org/wiki/Q844240","display_name":"Computer vision","level":1,"score":0.28290000557899475},{"id":"https://openalex.org/C33676613","wikidata":"https://www.wikidata.org/wiki/Q13415176","display_name":"Dimension (graph theory)","level":2,"score":0.27489998936653137},{"id":"https://openalex.org/C115961682","wikidata":"https://www.wikidata.org/wiki/Q860623","display_name":"Image (mathematics)","level":2,"score":0.2711000144481659},{"id":"https://openalex.org/C2779304628","wikidata":"https://www.wikidata.org/wiki/Q3503480","display_name":"Face (sociological concept)","level":2,"score":0.2709999978542328},{"id":"https://openalex.org/C52740198","wikidata":"https://www.wikidata.org/wiki/Q1539564","display_name":"Importance sampling","level":3,"score":0.26829999685287476},{"id":"https://openalex.org/C126255220","wikidata":"https://www.wikidata.org/wiki/Q141495","display_name":"Mathematical optimization","level":1,"score":0.2623000144958496},{"id":"https://openalex.org/C99498987","wikidata":"https://www.wikidata.org/wiki/Q2210247","display_name":"Noise (video)","level":3,"score":0.25609999895095825},{"id":"https://openalex.org/C159620131","wikidata":"https://www.wikidata.org/wiki/Q1938983","display_name":"Spatial analysis","level":2,"score":0.25099998712539673},{"id":"https://openalex.org/C106131492","wikidata":"https://www.wikidata.org/wiki/Q3072260","display_name":"Filter (signal processing)","level":2,"score":0.25}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2604.26503","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2604.26503","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"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":"doi:10.48550/arxiv.2604.26503","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2604.26503","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"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":false,"raw_source_name":null,"raw_type":"Preprint"},"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":{"Diffusion":[0],"models":[1,45],"have":[2],"achieved":[3],"remarkable":[4],"success":[5],"in":[6,26,46,71,184],"synthesizing":[7],"complex":[8],"static":[9],"and":[10,68,136,153,201,220],"temporal":[11,69,221],"visuals,":[12],"a":[13,37,47,100,117,130,151,166],"breakthrough":[14],"largely":[15],"driven":[16],"by":[17],"Classifier-Free":[18],"Guidance":[19,149],"(CFG).":[20],"However,":[21],"despite":[22],"its":[23],"pivotal":[24],"role":[25],"aligning":[27],"generated":[28],"content":[29],"with":[30],"textual":[31],"prompts,":[32],"standard":[33],"CFG":[34,97],"relies":[35],"on":[36,140],"globally":[38],"uniform":[39,113],"scalar.":[40],"This":[41],"homogeneous":[42],"amplification":[43],"traps":[44],"well-documented":[48],"\"detail-artifact":[49],"dilemma\":":[50],"low":[51],"guidance":[52,163],"scales":[53,61],"fail":[54],"to":[55,170,174,187],"inject":[56],"intricate":[57],"semantics,":[58],"while":[59,178],"high":[60],"inevitably":[62],"cause":[63],"structural":[64,218],"degradation,":[65],"color":[66],"over-saturation,":[67],"inconsistencies":[70],"videos.":[72],"In":[73],"this":[74,82,112],"paper,":[75],"we":[76,94,128,144],"expose":[77],"the":[78,85,105,123,211],"physical":[79],"root":[80],"of":[81,87],"flaw":[83],"through":[84],"lens":[86],"differential":[88],"geometry.":[89],"By":[90],"analyzing":[91],"Tweedie's":[92],"Formula,":[93],"reveal":[95],"that":[96,207],"intrinsically":[98],"performs":[99],"tangential":[101],"linear":[102,114],"extrapolation.":[103],"Because":[104],"natural":[106],"data":[107],"manifold":[108],"is":[109],"highly":[110],"curved,":[111],"step":[115],"introduces":[116],"severe":[118],"orthogonal":[119],"deviation.":[120],"To":[121],"keep":[122],"generation":[124],"trajectory":[125],"safely":[126],"bounded,":[127],"formulate":[129],"theoretical":[131],"upper":[132],"bound":[133],"for":[134],"spatial":[135],"adaptive":[137],"guidance.":[138],"Based":[139],"these":[141],"geometric":[142],"insights,":[143],"propose":[145],"Spatial":[146],"Adaptive":[147],"Multi":[148],"(SAMG),":[150],"training-free":[152],"virtually":[154],"zero-cost":[155],"sampling":[156],"algorithm.":[157],"SAMG":[158,208],"dynamically":[159],"computes":[160],"point-wise":[161],"conditional":[162],"energy,":[164],"applying":[165],"conservative":[167],"minimum":[168],"scale":[169,183],"high-energy":[171],"boundary":[172],"regions":[173,186],"preserve":[175],"delicate":[176],"micro-textures,":[177],"deploying":[179],"an":[180],"aggressive":[181],"maximum":[182],"low-energy":[185],"maximize":[188],"semantic":[189,216],"injection.":[190],"Extensive":[191],"experiments":[192],"across":[193],"diverse":[194],"image":[195],"(SD":[196],"1.5,":[197],"SDXL,":[198],"SD3.5":[199],"Medium)":[200],"video":[202],"(CogVideoX,":[203],"ModelScope)":[204],"architectures":[205],"demonstrate":[206],"effectively":[209],"resolves":[210],"detail-artifact":[212],"dilemma,":[213],"achieving":[214],"superior":[215],"alignment,":[217],"integrity,":[219],"smoothness":[222],"without":[223],"any":[224],"computational":[225],"overhead.":[226]},"counts_by_year":[],"updated_date":"2026-07-01T06:00:48.157686","created_date":"2026-05-01T00:00:00"}