{"id":"https://openalex.org/W7214767013","doi":"https://doi.org/10.48550/arxiv.2609.34371","title":"From Static to Dynamic: On-Policy Distillation from Image to Video Diffusion Models","display_name":"From Static to Dynamic: On-Policy Distillation from Image to Video Diffusion Models","publication_year":2026,"publication_date":"2026-09-28","ids":{"openalex":"https://openalex.org/W7214767013","doi":"https://doi.org/10.48550/arxiv.2609.34371"},"language":"en","primary_location":{"id":"pmh:oai:arXiv.org:2609.34371","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2609.34371","pdf_url":"https://arxiv.org/pdf/2609.34371","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/2609.34371","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5135876135","display_name":"Bingqing Jiang","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I889458895","display_name":"University of Hong Kong","ror":"https://ror.org/02zhqgq86","country_code":"HK","type":"education","lineage":["https://openalex.org/I889458895"]}],"countries":["HK"],"is_corresponding":false,"raw_author_name":"Jiang, Bingqing","raw_affiliation_strings":["The University of Hong Kong"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"The University of Hong Kong","institution_ids":["https://openalex.org/I889458895"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5153455249","display_name":"Li Luo","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I889458895","display_name":"University of Hong Kong","ror":"https://ror.org/02zhqgq86","country_code":"HK","type":"education","lineage":["https://openalex.org/I889458895"]}],"countries":["HK"],"is_corresponding":false,"raw_author_name":"Luo, Li","raw_affiliation_strings":["The University of Hong Kong"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"The University of Hong Kong","institution_ids":["https://openalex.org/I889458895"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5112840113","display_name":"Zichao Yu","orcid":"https://orcid.org/0009-0005-8880-279X","observed_orcids":["https://orcid.org/0009-0005-8880-279X"]},"institutions":[{"id":"https://openalex.org/I889458895","display_name":"University of Hong Kong","ror":"https://ror.org/02zhqgq86","country_code":"HK","type":"education","lineage":["https://openalex.org/I889458895"]}],"countries":["HK"],"is_corresponding":false,"raw_author_name":"Yu, Zichao","raw_affiliation_strings":["The University of Hong Kong"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"The University of Hong Kong","institution_ids":["https://openalex.org/I889458895"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5153421920","display_name":"Yujin Han","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I889458895","display_name":"University of Hong Kong","ror":"https://ror.org/02zhqgq86","country_code":"HK","type":"education","lineage":["https://openalex.org/I889458895"]}],"countries":["HK"],"is_corresponding":false,"raw_author_name":"Han, Yujin","raw_affiliation_strings":["The University of Hong Kong"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"The University of Hong Kong","institution_ids":["https://openalex.org/I889458895"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5153479211","display_name":"Zhaolong Su","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I205783295","display_name":"Cornell University","ror":"https://ror.org/05bnh6r87","country_code":"US","type":"education","lineage":["https://openalex.org/I205783295"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Su, Zhaolong","raw_affiliation_strings":["Cornell University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Cornell University","institution_ids":["https://openalex.org/I205783295"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5085848346","display_name":"Difan Zou","orcid":"https://orcid.org/0000-0002-6544-2593","observed_orcids":["https://orcid.org/0000-0002-6544-2593"]},"institutions":[{"id":"https://openalex.org/I889458895","display_name":"University of Hong Kong","ror":"https://ror.org/02zhqgq86","country_code":"HK","type":"education","lineage":["https://openalex.org/I889458895"]}],"countries":["HK"],"is_corresponding":false,"raw_author_name":"Zou, Difan","raw_affiliation_strings":["The University of Hong Kong"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"The University of Hong Kong","institution_ids":["https://openalex.org/I889458895"]}]}],"institutions":[],"countries_distinct_count":2,"institutions_distinct_count":2,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":true,"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":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/T11714","display_name":"Multimodal Machine Learning Applications","score":0.0,"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/T12357","display_name":"Digital Media Forensic Detection","score":0.0,"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"}}],"keywords":[{"id":"https://openalex.org/keywords/on-policy-distillation","display_name":"on-policy distillation","score":0.9980000257492065},{"id":"https://openalex.org/keywords/video-diffusion-models","display_name":"video diffusion models","score":0.781000018119812},{"id":"https://openalex.org/keywords/latent-space-alignment","display_name":"latent space alignment","score":0.7049999833106995},{"id":"https://openalex.org/keywords/motion-preservation","display_name":"motion preservation","score":0.628000020980835},{"id":"https://openalex.org/keywords/temporal-coherence","display_name":"temporal coherence","score":0.9629999995231628},{"id":"https://openalex.org/keywords/video-generation","display_name":"video generation","score":0.5080000162124634}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7732999920845032},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5435000061988831},{"id":"https://openalex.org/C115961682","wikidata":"https://www.wikidata.org/wiki/Q860623","display_name":"Image (mathematics)","level":2,"score":0.5347999930381775},{"id":"https://openalex.org/C31972630","wikidata":"https://www.wikidata.org/wiki/Q844240","display_name":"Computer vision","level":1,"score":0.5098999738693237},{"id":"https://openalex.org/C2776436953","wikidata":"https://www.wikidata.org/wiki/Q5163215","display_name":"Consistency (knowledge bases)","level":2,"score":0.4984000027179718},{"id":"https://openalex.org/C104114177","wikidata":"https://www.wikidata.org/wiki/Q79782","display_name":"Motion (physics)","level":2,"score":0.4643000066280365},{"id":"https://openalex.org/C75553542","wikidata":"https://www.wikidata.org/wiki/Q178161","display_name":"A priori and a posteriori","level":2,"score":0.4117000102996826},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.4108000099658966},{"id":"https://openalex.org/C204030448","wikidata":"https://www.wikidata.org/wiki/Q101017","display_name":"Distillation","level":2,"score":0.35100001096725464},{"id":"https://openalex.org/C65483669","wikidata":"https://www.wikidata.org/wiki/Q3536669","display_name":"Video processing","level":2,"score":0.32829999923706055},{"id":"https://openalex.org/C9417928","wikidata":"https://www.wikidata.org/wiki/Q1070689","display_name":"Image processing","level":3,"score":0.3257000148296356},{"id":"https://openalex.org/C55020928","wikidata":"https://www.wikidata.org/wiki/Q3813865","display_name":"Image quality","level":3,"score":0.32089999318122864},{"id":"https://openalex.org/C165064840","wikidata":"https://www.wikidata.org/wiki/Q1321061","display_name":"Matching (statistics)","level":2,"score":0.30640000104904175},{"id":"https://openalex.org/C2779530757","wikidata":"https://www.wikidata.org/wiki/Q1207505","display_name":"Quality (philosophy)","level":2,"score":0.3041999936103821},{"id":"https://openalex.org/C103910844","wikidata":"https://www.wikidata.org/wiki/Q2631256","display_name":"Video quality","level":3,"score":0.27880001068115234},{"id":"https://openalex.org/C10161872","wikidata":"https://www.wikidata.org/wiki/Q557891","display_name":"Motion estimation","level":2,"score":0.26910001039505005},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.2567000091075897},{"id":"https://openalex.org/C2776175482","wikidata":"https://www.wikidata.org/wiki/Q1195816","display_name":"Transfer (computing)","level":2,"score":0.25209999084472656},{"id":"https://openalex.org/C153180895","wikidata":"https://www.wikidata.org/wiki/Q7148389","display_name":"Pattern recognition (psychology)","level":2,"score":0.251800000667572}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:arXiv.org:2609.34371","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2609.34371","pdf_url":"https://arxiv.org/pdf/2609.34371","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.2609.34371","doi":"https://doi.org/10.48550/arxiv.2609.34371","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2609.34371","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:2609.34371","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2609.34371","pdf_url":"https://arxiv.org/pdf/2609.34371","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":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W7214767013.pdf","grobid_xml":"https://content.openalex.org/works/W7214767013.grobid-xml"},"referenced_works_count":15,"referenced_works":["https://openalex.org/W2908510526","https://openalex.org/W3168867926","https://openalex.org/W4298185919","https://openalex.org/W4383994075","https://openalex.org/W4400612128","https://openalex.org/W4405253641","https://openalex.org/W4406052102","https://openalex.org/W4409092337","https://openalex.org/W4415237773","https://openalex.org/W4415566463","https://openalex.org/W4416780405","https://openalex.org/W7159641083","https://openalex.org/W7160838488","https://openalex.org/W7161243847","https://openalex.org/W7202205122"],"related_works":["https://openalex.org/W4415795783","https://openalex.org/W7124435510","https://openalex.org/W4393027525","https://openalex.org/W7165686740","https://openalex.org/W7162764283","https://openalex.org/W7134916327","https://openalex.org/W7220462274","https://openalex.org/W7140183588","https://openalex.org/W7171723646","https://openalex.org/W6929537309"],"abstract_inverted_index":{"On-policy":[0],"distillation":[1,207],"(OPD)":[2],"specializes":[3],"pretrained":[4,120,156],"video":[5,19,26,31,77,121,161,214],"diffusion":[6],"models":[7,20],"through":[8],"teacher":[9],"supervision":[10,82,143],"along":[11],"the":[12,155,219,225],"student's":[13,157],"own":[14],"generation":[15,215],"trajectory.":[16],"Although":[17],"large":[18],"are":[21],"natural":[22],"teachers,":[23],"developing":[24],"specialized":[25,115,177],"experts":[27,54,89,179],"can":[28],"require":[29],"costly":[30],"data":[32],"and":[33,49,67,76,134,163,173,180,200,221],"training,":[34],"while":[35,87,118],"querying":[36,43],"them":[37],"incurs":[38],"substantially":[39],"higher":[40],"latency":[41],"than":[42],"image":[44,53,75,88,116,140,178],"experts.":[45],"More":[46],"readily":[47],"available":[48],"cheaper":[50],"to":[51,98,159,169,212],"query,":[52],"offer":[55],"a":[56,107,125],"cost-effective":[57],"alternative,":[58],"particularly":[59],"for":[60],"largely":[61],"temporal-agnostic":[62],"capabilities":[63,223],"such":[64],"as":[65,208],"aesthetics":[66],"OCR":[68],"that":[69,113,129],"admit":[70],"frame-level":[71,99],"supervision.":[72],"However,":[73],"heterogeneous":[74,146,201],"latent":[78,132,147],"spaces":[79],"prevent":[80],"direct":[81],"of":[83,139,224],"intermediate":[84],"student":[85,131],"states,":[86],"lack":[90],"cross-frame":[91],"motion":[92,167,171],"supervision,":[93],"making":[94],"temporal":[95,174],"consistency":[96],"vulnerable":[97],"improvements.":[100],"In":[101],"this":[102],"paper,":[103],"we":[104],"propose":[105],"MILD,":[106],"Motion-Preserving":[108],"Image-to-Video":[109],"Latent":[110],"Distillation":[111],"framework":[112],"transfers":[114],"expertise":[117],"preserving":[119],"dynamics.":[122],"MILD":[123],"uses":[124],"learnable":[126],"linear":[127],"connector":[128,198],"aligns":[130],"states":[133],"predicted":[135],"updates":[136],"with":[137,191],"those":[138],"experts,":[141],"enabling":[142],"transfer":[144,196],"across":[145,197],"spaces.":[148],"We":[149],"further":[150,192],"constrain":[151],"image-guided":[152],"corrections":[153],"around":[154],"predictions":[158],"preserve":[160],"dynamics":[162],"incorporate":[164],"an":[165,209],"optical-flow-based":[166],"reward":[168],"improve":[170],"quality":[172],"consistency.":[175],"Across":[176],"multiple":[181],"video-student":[182],"backbones,":[183],"our":[184],"method":[185],"consistently":[186],"outperforms":[187],"video-teacher":[188],"OPD":[189],"baselines,":[190],"studies":[193],"demonstrating":[194],"effective":[195,210],"designs":[199],"architectures.":[202],"These":[203],"results":[204],"establish":[205],"image-to-video":[206],"route":[211],"improving":[213],"by":[216],"drawing":[217],"on":[218],"diverse":[220],"evolving":[222],"image-generation":[226],"ecosystem.":[227]},"counts_by_year":[],"updated_date":"2026-11-07T11:25:07.637180","created_date":"2026-09-30T00:00:00"}