{"id":1179134,"date":"2026-07-20T06:41:48","date_gmt":"2026-07-20T13:41:48","guid":{"rendered":"https:\/\/find.codeghost.online\/en-us\/research\/publication\/combostoc-combinatorial-stochasticity-for-diffusion-generative-models\/"},"modified":"2026-07-21T13:33:23","modified_gmt":"2026-07-21T20:33:23","slug":"combostoc-combinatorial-stochasticity-for-diffusion-generative-models","status":"publish","type":"msr-research-item","link":"https:\/\/find.codeghost.online\/en-us\/research\/publication\/combostoc-combinatorial-stochasticity-for-diffusion-generative-models\/","title":{"rendered":"ComboStoc: Combinatorial Stochasticity for Diffusion Generative Models"},"content":{"rendered":"\n\n\n<p class=\"wp-block-paragraph\">In this paper, we study an under-explored but important factor of diffusion generative models, i.e., the combinatorial complexity. Data samples are generally high-dimensional, and for various structured generation tasks, additional attributes are combined to associate with data samples. We show that the space spanned by the combination of dimensions and attributes can be insufficiently covered by existing training schemes of diffusion generative models, potentially limiting test time performance. We present a simple fix to this problem by constructing stochastic processes that fully exploit the combinatorial structures, hence the name ComboStoc. Using this simple strategy, we show that network training is significantly accelerated across diverse data modalities, including images and 3D structured shapes. Moreover, ComboStoc enables a new way of test time generation which uses asynchronous time steps for different dimensions and attributes, thus allowing for varying degrees of control over them. Our code is available at: <a class=\"msr-external-link glyph-append glyph-append-open-in-new-tab glyph-append-xsmall\" href=\"https:\/\/github.com\/Xrvitd\/ComboStoc\" target=\"_blank\" rel=\"noopener noreferrer\">https:\/\/github.com\/Xrvitd\/ComboStoc<span class=\"sr-only\"> (opens in new tab)<\/span><\/a>.<\/p>\n","protected":false},"excerpt":{"rendered":"<p>In this paper, we study an under-explored but important factor of diffusion generative models, i.e., the combinatorial complexity. Data samples are generally high-dimensional, and for various structured generation tasks, additional attributes are combined to associate with data samples. We show that the space spanned by the combination of dimensions and attributes can be insufficiently covered [&hellip;]<\/p>\n","protected":false},"featured_media":0,"template":"","meta":{"msr-url-field":"","msr-podcast-episode":"","msrModifiedDate":"","msrModifiedDateEnabled":false,"ep_exclude_from_search":false,"_classifai_error":"","msr-author-ordering":[{"type":"text","value":"Rui Xu","user_id":0},{"type":"text","value":"Jie-Chao Wang","user_id":0},{"type":"user_nicename","value":"Hao Pan","user_id":"31970"},{"type":"user_nicename","value":"Yang Liu","user_id":"39594"},{"type":"user_nicename","value":"Xin Tong","user_id":"34929"},{"type":"text","value":"Shiqing Xin","user_id":0},{"type":"text","value":"Changhe Tu","user_id":0},{"type":"text","value":"Taku Komura","user_id":0},{"type":"text","value":"Wenping Wang","user_id":0}],"msr_publishername":"","msr_publisher_other":"","msr_booktitle":"","msr_chapter":"","msr_edition":"","msr_editors":"","msr_how_published":"","msr_isbn":"","msr_issue":"","msr_journal":"","msr_number":"","msr_organization":"","msr_pages_string":"","msr_page_range_start":"","msr_page_range_end":"","msr_series":"","msr_volume":"","msr_copyright":"","msr_conference_name":"SIGGRAPH 2026","msr_doi":"10.1145\/3811285","msr_arxiv_id":"2405.13729","msr_mag_id":"","msr_other_authors":"","msr_other_contributors":"","msr_speaker":"","msr_award":"","msr_affiliation":"","msr_institution":"","msr_host":"","msr_version":"","msr_duration":"","msr_release_tracker_id":"","msr_highlight_type":"","msr_date_display_format":"","msr_main_download_label":"","msr_external_link_label":"","msr_doi_label":"","msr_published_date":"2026-04-29","msr_startdate":"","msr_presentation_date":"","msr_highlight_text":"","msr_notes":"","msr_longbiography":"","msr_publicationurl":"","msr_external_url":"","msr_secondary_video_url":"","msr_conference_url":"","msr_journal_url":"","msr_year":2026,"msr_month":4,"msr_day":29,"msr_microsoftintellectualproperty":false,"msr_pub_id":"f7a2e69fc89e42dcf48a0320b453ccc2bdc3ee7a","msr_publication_uploader":[{"type":"url","title":"https:\/\/dblp.org\/rec\/journals\/corr\/abs-2405-13729.html","label_id":243109,"id":false,"viewUrl":false},{"type":"url","title":"https:\/\/arxiv.org\/abs\/2405.13729","label_id":243109,"id":false,"viewUrl":false}],"msr_related_uploader":[],"msr_original_fields_of_study":[],"msr_s2_paper_id":"f7a2e69fc89e42dcf48a0320b453ccc2bdc3ee7a","msr_s2_pdf_url":"https:\/\/doi.org\/10.1145\/3811285","msr_citation_count_updated":"","msr_citation_count":3,"msr_influential_citations":0,"msr_reference_count":57,"msr_s2_open_access":true,"msr_s2_author_ids":[],"msr_pub_ids":[{"provider":"s2","id":"f7a2e69fc89e42dcf48a0320b453ccc2bdc3ee7a"},{"provider":"doi","id":"10.1145\/3811285"},{"provider":"arxiv","id":"2405.13729"},{"provider":"dblp","id":"journals\/corr\/abs-2405-13729"},{"provider":"corpusid","id":"269983533"}],"msr_hide_image_in_river":0,"footnotes":""},"msr-research-highlight":[],"research-area":[13556,13562],"msr-publication-type":[193716],"msr-publisher":[],"msr-publication-cta":[],"msr-focus-area":[],"msr-locale":[268875],"msr-post-option":[],"msr-field-of-study":[246691,263185,265497],"msr-conference":[],"msr-journal":[270030],"msr-impact-theme":[],"msr-pillar":[],"class_list":["post-1179134","msr-research-item","type-msr-research-item","status-publish","hentry","msr-research-area-artificial-intelligence","msr-research-area-computer-vision","msr-locale-en_us","msr-field-of-study-computer-science","msr-field-of-study-computer-vision-and-pattern-recognition","msr-field-of-study-machine-learning-296"],"msr_publishername":"","msr_edition":"","msr_affiliation":"","msr_published_date":"2026-04-29","msr_host":"","msr_duration":"","msr_version":"","msr_speaker":"","msr_other_contributors":"","msr_booktitle":"","msr_pages_string":"","msr_chapter":"","msr_isbn":"","msr_journal":"","msr_volume":"","msr_number":"","msr_editors":"","msr_series":"","msr_issue":"","msr_organization":"","msr_how_published":"","msr_notes":"","msr_highlight_text":"","msr_release_tracker_id":"","msr_original_fields_of_study":"","msr_download_urls":"","msr_external_url":"","msr_secondary_video_url":"","msr_longbiography":"","msr_microsoftintellectualproperty":0,"msr_main_download":"","msr_publicationurl":"","msr_doi":"10.1145\/3811285","msr_publication_uploader":[{"type":"url","title":"https:\/\/dblp.org\/rec\/journals\/corr\/abs-2405-13729.html","label_id":243109,"id":false,"viewUrl":false},{"type":"url","title":"https:\/\/arxiv.org\/abs\/2405.13729","label_id":243109,"id":false,"viewUrl":false}],"msr_related_uploader":[],"msr_citation_count":3,"msr_citation_count_updated":"","msr_s2_paper_id":"f7a2e69fc89e42dcf48a0320b453ccc2bdc3ee7a","msr_influential_citations":0,"msr_reference_count":57,"msr_arxiv_id":"2405.13729","msr_s2_author_ids":[],"msr_s2_open_access":true,"msr_s2_pdf_url":"https:\/\/doi.org\/10.1145\/3811285","msr_attachments":[],"msr-author-ordering":[{"type":"text","value":"Rui Xu","user_id":0,"rest_url":false},{"type":"text","value":"Jie-Chao Wang","user_id":0,"rest_url":false},{"type":"user_nicename","value":"Hao Pan","user_id":31970,"rest_url":"https:\/\/find.codeghost.online\/en-us\/research\/wp-json\/microsoft-research\/v1\/researchers?person=Hao Pan"},{"type":"user_nicename","value":"Yang Liu","user_id":39594,"rest_url":"https:\/\/find.codeghost.online\/en-us\/research\/wp-json\/microsoft-research\/v1\/researchers?person=Yang Liu"},{"type":"user_nicename","value":"Xin Tong","user_id":34929,"rest_url":"https:\/\/find.codeghost.online\/en-us\/research\/wp-json\/microsoft-research\/v1\/researchers?person=Xin Tong"},{"type":"text","value":"Shiqing Xin","user_id":0,"rest_url":false},{"type":"text","value":"Changhe Tu","user_id":0,"rest_url":false},{"type":"text","value":"Taku Komura","user_id":0,"rest_url":false},{"type":"text","value":"Wenping Wang","user_id":0,"rest_url":false}],"msr_impact_theme":[],"msr_research_lab":[199560],"msr_event":[],"msr_group":[],"msr_project":[],"publication":[],"video":[],"msr-tool":[],"msr_publication_type":"inproceedings","related_content":[],"_links":{"self":[{"href":"https:\/\/find.codeghost.online\/en-us\/research\/wp-json\/wp\/v2\/msr-research-item\/1179134","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/find.codeghost.online\/en-us\/research\/wp-json\/wp\/v2\/msr-research-item"}],"about":[{"href":"https:\/\/find.codeghost.online\/en-us\/research\/wp-json\/wp\/v2\/types\/msr-research-item"}],"version-history":[{"count":2,"href":"https:\/\/find.codeghost.online\/en-us\/research\/wp-json\/wp\/v2\/msr-research-item\/1179134\/revisions"}],"predecessor-version":[{"id":1179443,"href":"https:\/\/find.codeghost.online\/en-us\/research\/wp-json\/wp\/v2\/msr-research-item\/1179134\/revisions\/1179443"}],"wp:attachment":[{"href":"https:\/\/find.codeghost.online\/en-us\/research\/wp-json\/wp\/v2\/media?parent=1179134"}],"wp:term":[{"taxonomy":"msr-research-highlight","embeddable":true,"href":"https:\/\/find.codeghost.online\/en-us\/research\/wp-json\/wp\/v2\/msr-research-highlight?post=1179134"},{"taxonomy":"msr-research-area","embeddable":true,"href":"https:\/\/find.codeghost.online\/en-us\/research\/wp-json\/wp\/v2\/research-area?post=1179134"},{"taxonomy":"msr-publication-type","embeddable":true,"href":"https:\/\/find.codeghost.online\/en-us\/research\/wp-json\/wp\/v2\/msr-publication-type?post=1179134"},{"taxonomy":"msr-publisher","embeddable":true,"href":"https:\/\/find.codeghost.online\/en-us\/research\/wp-json\/wp\/v2\/msr-publisher?post=1179134"},{"taxonomy":"msr-publication-cta","embeddable":true,"href":"https:\/\/find.codeghost.online\/en-us\/research\/wp-json\/wp\/v2\/msr-publication-cta?post=1179134"},{"taxonomy":"msr-focus-area","embeddable":true,"href":"https:\/\/find.codeghost.online\/en-us\/research\/wp-json\/wp\/v2\/msr-focus-area?post=1179134"},{"taxonomy":"msr-locale","embeddable":true,"href":"https:\/\/find.codeghost.online\/en-us\/research\/wp-json\/wp\/v2\/msr-locale?post=1179134"},{"taxonomy":"msr-post-option","embeddable":true,"href":"https:\/\/find.codeghost.online\/en-us\/research\/wp-json\/wp\/v2\/msr-post-option?post=1179134"},{"taxonomy":"msr-field-of-study","embeddable":true,"href":"https:\/\/find.codeghost.online\/en-us\/research\/wp-json\/wp\/v2\/msr-field-of-study?post=1179134"},{"taxonomy":"msr-conference","embeddable":true,"href":"https:\/\/find.codeghost.online\/en-us\/research\/wp-json\/wp\/v2\/msr-conference?post=1179134"},{"taxonomy":"msr-journal","embeddable":true,"href":"https:\/\/find.codeghost.online\/en-us\/research\/wp-json\/wp\/v2\/msr-journal?post=1179134"},{"taxonomy":"msr-impact-theme","embeddable":true,"href":"https:\/\/find.codeghost.online\/en-us\/research\/wp-json\/wp\/v2\/msr-impact-theme?post=1179134"},{"taxonomy":"msr-pillar","embeddable":true,"href":"https:\/\/find.codeghost.online\/en-us\/research\/wp-json\/wp\/v2\/msr-pillar?post=1179134"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}