{"id":"https://openalex.org/W7221010232","doi":"https://doi.org/10.1109/cvprf71909.2026.00895","title":"DocSLM: A Small Vision-Language Model for Long Multimodal Document Understanding","display_name":"DocSLM: A Small Vision-Language Model for Long Multimodal Document Understanding","publication_year":2026,"publication_date":"2026-06-01","ids":{"openalex":"https://openalex.org/W7221010232","doi":"https://doi.org/10.1109/cvprf71909.2026.00895"},"language":"en","primary_location":{"id":"doi:10.1109/cvprf71909.2026.00895","doi":"https://doi.org/10.1109/cvprf71909.2026.00895","is_oa":true,"landing_page_url":"https://doi.org/10.1109/cvprf71909.2026.00895","pdf_url":null,"source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2026 IEEE/CVF Conference on Computer Vision and Pattern Recognition- FINDINGS Track (CVPRF)","raw_type":"proceedings-article"},"volume":null,"type":"conference-paper","is_traditional":true,"indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://doi.org/10.1109/cvprf71909.2026.00895","any_repository_has_fulltext":null},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5003653508","display_name":"Tanveer Hannan","orcid":"https://orcid.org/0000-0001-9957-9531","observed_orcids":["https://orcid.org/0000-0001-9957-9531"]},"institutions":[{"id":"https://openalex.org/I8204097","display_name":"Ludwig-Maximilians-Universit\u00e4t M\u00fcnchen","ror":"https://ror.org/05591te55","country_code":"DE","type":"education","lineage":["https://openalex.org/I8204097"]}],"countries":["DE"],"is_corresponding":false,"raw_author_name":"Tanveer Hannan","raw_affiliation_strings":["LMU Munich"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"LMU Munich","institution_ids":["https://openalex.org/I8204097"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5159286593","display_name":"Dimitrios Mallios","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I1290206253","display_name":"Microsoft (United States)","ror":"https://ror.org/00d0nc645","country_code":"US","type":"company","lineage":["https://openalex.org/I1290206253"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Dimitrios Mallios","raw_affiliation_strings":["Microsoft"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Microsoft","institution_ids":["https://openalex.org/I1290206253"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5138514556","display_name":"Parth Pathak","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Parth Pathak","raw_affiliation_strings":["FAIR Meta"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"FAIR Meta","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5017782637","display_name":"Faegheh Sardari","orcid":"https://orcid.org/0000-0002-9134-0427","observed_orcids":["https://orcid.org/0000-0002-9134-0427"]},"institutions":[{"id":"https://openalex.org/I1290206253","display_name":"Microsoft (United States)","ror":"https://ror.org/00d0nc645","country_code":"US","type":"company","lineage":["https://openalex.org/I1290206253"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Faegheh Sardari","raw_affiliation_strings":["Microsoft"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Microsoft","institution_ids":["https://openalex.org/I1290206253"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5003335849","display_name":"Thomas Seidl","orcid":"https://orcid.org/0000-0002-4861-1412","observed_orcids":["https://orcid.org/0000-0002-4861-1412"]},"institutions":[{"id":"https://openalex.org/I8204097","display_name":"Ludwig-Maximilians-Universit\u00e4t M\u00fcnchen","ror":"https://ror.org/05591te55","country_code":"DE","type":"education","lineage":["https://openalex.org/I8204097"]}],"countries":["DE"],"is_corresponding":false,"raw_author_name":"Thomas Seidl","raw_affiliation_strings":["LMU Munich"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"LMU Munich","institution_ids":["https://openalex.org/I8204097"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5081800468","display_name":"Gedas Bertasius","orcid":"https://orcid.org/0000-0003-1800-4790","observed_orcids":["https://orcid.org/0000-0003-1800-4790"]},"institutions":[{"id":"https://openalex.org/I114027177","display_name":"University of North Carolina at Chapel Hill","ror":"https://ror.org/0130frc33","country_code":"US","type":"education","lineage":["https://openalex.org/I114027177"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Gedas Bertasius","raw_affiliation_strings":["UNC Chapel Hill"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"UNC Chapel Hill","institution_ids":["https://openalex.org/I114027177"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5031986521","display_name":"Mohsen Fayyaz","orcid":"https://orcid.org/0000-0001-5254-3128","observed_orcids":["https://orcid.org/0000-0001-5254-3128"]},"institutions":[{"id":"https://openalex.org/I1290206253","display_name":"Microsoft (United States)","ror":"https://ror.org/00d0nc645","country_code":"US","type":"company","lineage":["https://openalex.org/I1290206253"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Mohsen Fayyaz","raw_affiliation_strings":["Microsoft"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Microsoft","institution_ids":["https://openalex.org/I1290206253"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5069905715","display_name":"Sunando Sengupta","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I1290206253","display_name":"Microsoft (United States)","ror":"https://ror.org/00d0nc645","country_code":"US","type":"company","lineage":["https://openalex.org/I1290206253"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Sunando Sengupta","raw_affiliation_strings":["Microsoft"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Microsoft","institution_ids":["https://openalex.org/I1290206253"]}]}],"institutions":[],"countries_distinct_count":2,"institutions_distinct_count":3,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":0.0,"has_fulltext":false,"cited_by_count":0,"citation_normalized_percentile":{"value":0.0,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"9338","last_page":"9347"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T11714","display_name":"Multimodal Machine Learning Applications","score":0.9945999979972839,"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/T11714","display_name":"Multimodal Machine Learning Applications","score":0.9945999979972839,"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/T10601","display_name":"Handwritten Text Recognition Techniques","score":0.005100000184029341,"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/T10036","display_name":"Advanced Neural Network Applications","score":9.999999747378752e-05,"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/multimodal-document-understanding","display_name":"multimodal document understanding","score":0.8960000276565552},{"id":"https://openalex.org/keywords/memory-efficient-models","display_name":"memory-efficient models","score":0.6129999756813049},{"id":"https://openalex.org/keywords/edge-devices","display_name":"edge devices","score":0.7440000176429749},{"id":"https://openalex.org/keywords/uncertainty-calibration","display_name":"uncertainty calibration","score":0.8849999904632568}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6100999712944031},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.3806000053882599},{"id":"https://openalex.org/C2776401178","wikidata":"https://www.wikidata.org/wiki/Q12050496","display_name":"Feature (linguistics)","level":2,"score":0.2937999963760376},{"id":"https://openalex.org/C9652623","wikidata":"https://www.wikidata.org/wiki/Q190109","display_name":"Field (mathematics)","level":2,"score":0.27469998598098755},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.27149999141693115},{"id":"https://openalex.org/C177264268","wikidata":"https://www.wikidata.org/wiki/Q1514741","display_name":"Set (abstract data type)","level":2,"score":0.25360000133514404},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.24699999392032623},{"id":"https://openalex.org/C192209626","wikidata":"https://www.wikidata.org/wiki/Q190909","display_name":"Focus (optics)","level":2,"score":0.24230000376701355},{"id":"https://openalex.org/C98045186","wikidata":"https://www.wikidata.org/wiki/Q205663","display_name":"Process (computing)","level":2,"score":0.2305999994277954},{"id":"https://openalex.org/C116834253","wikidata":"https://www.wikidata.org/wiki/Q2039217","display_name":"Identification (biology)","level":2,"score":0.22920000553131104}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/cvprf71909.2026.00895","doi":"https://doi.org/10.1109/cvprf71909.2026.00895","is_oa":true,"landing_page_url":"https://doi.org/10.1109/cvprf71909.2026.00895","pdf_url":null,"source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2026 IEEE/CVF Conference on Computer Vision and Pattern Recognition- FINDINGS Track (CVPRF)","raw_type":"proceedings-article"}],"best_oa_location":{"id":"doi:10.1109/cvprf71909.2026.00895","doi":"https://doi.org/10.1109/cvprf71909.2026.00895","is_oa":true,"landing_page_url":"https://doi.org/10.1109/cvprf71909.2026.00895","pdf_url":null,"source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2026 IEEE/CVF Conference on Computer Vision and Pattern Recognition- FINDINGS Track (CVPRF)","raw_type":"proceedings-article"},"sustainable_development_goals":[],"study_designs":[],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":26,"referenced_works":["https://openalex.org/W2947555604","https://openalex.org/W4312263373","https://openalex.org/W4319300063","https://openalex.org/W4386075667","https://openalex.org/W4390873674","https://openalex.org/W4390873989","https://openalex.org/W4402671775","https://openalex.org/W4402716395","https://openalex.org/W4404782234","https://openalex.org/W4405595839","https://openalex.org/W4411725552","https://openalex.org/W4412945734","https://openalex.org/W4413145389","https://openalex.org/W4413146357","https://openalex.org/W4413146441","https://openalex.org/W4413146877","https://openalex.org/W4413147436","https://openalex.org/W4413259061","https://openalex.org/W4414197839","https://openalex.org/W4415796209","https://openalex.org/W4415796627","https://openalex.org/W4415798827","https://openalex.org/W4416167273","https://openalex.org/W7133196460","https://openalex.org/W7133227460","https://openalex.org/W7160034663"],"related_works":["https://openalex.org/W4416036830","https://openalex.org/W4402671775","https://openalex.org/W7214577769","https://openalex.org/W3176851559","https://openalex.org/W7220987193","https://openalex.org/W4404783389","https://openalex.org/W4412887059","https://openalex.org/W4385571998","https://openalex.org/W7155559420","https://openalex.org/W4411119769"],"abstract_inverted_index":{"Large":[0],"Vision-Language":[1,35],"Models":[2],"(LVLMs)":[3],"have":[4],"demonstrated":[5],"strong":[6],"multimodal":[7,111,134],"reasoning":[8],"capabilities":[9],"on":[10,25,95,137],"long":[11,39,84,110],"and":[12,57,75,99,129],"complex":[13],"documents.":[14],"However,":[15],"their":[16],"high":[17],"memory":[18,44,69],"footprint":[19],"makes":[20],"them":[21],"impractical":[22],"for":[23,38],"deployment":[24],"resourceconstrained":[26],"edge":[27,139],"devices.":[28,140],"We":[29],"present":[30],"DocSLM,":[31],"an":[32,104],"efficient":[33],"Small":[34],"Model":[36],"designed":[37],"document":[40,96,112,135],"understanding":[41,136],"under":[42],"constrained":[43],"resources.":[45],"DocSLM":[46,114],"incorporates":[47],"a":[48,64,89],"Hierarchical":[49],"Multimodal":[50],"Compressor":[51],"that":[52,93],"jointly":[53],"encodes":[54],"visual,":[55],"textual,":[56],"layout":[58],"information":[59],"from":[60],"each":[61],"page":[62],"into":[63],"fixed-length":[65],"sequence,":[66],"greatly":[67],"reducing":[68],"consumption":[70],"while":[71,120],"preserving":[72],"both":[73],"local":[74],"global":[76],"semantics.":[77],"To":[78],"enable":[79],"scalable":[80],"processing":[81],"over":[82],"arbitrarily":[83],"inputs,":[85],"we":[86],"further":[87],"introduce":[88],"Streaming":[90],"Abstention":[91],"mechanism":[92],"operates":[94],"segments":[97],"sequentially":[98],"filters":[100],"low-confidence":[101],"responses":[102],"through":[103],"entropy-based":[105],"uncertainty":[106],"calibrator.":[107],"Across":[108],"multiple":[109],"benchmarks,":[113],"matches":[115],"or":[116],"surpasses":[117],"state-of-the-art":[118],"methods":[119],"using":[121],"82%":[122],"fewer":[123,127],"visual":[124],"tokens,":[125],"75%":[126],"parameters,":[128],"71%":[130],"lower":[131],"latency-delivering":[132],"reliable":[133],"lightweight":[138],"Code":[141],"is":[142],"available":[143],"in":[144],"the":[145],"supplementary.":[146]},"counts_by_year":[],"updated_date":"2026-11-08T12:12:28.789561","created_date":"2026-10-09T00:00:00"}