{"id":"https://openalex.org/W4402727885","doi":"https://doi.org/10.1109/cvpr52733.2024.02520","title":"VILA: On Pre-training for Visual Language Models","display_name":"VILA: On Pre-training for Visual Language Models","publication_year":2024,"publication_date":"2024-06-16","ids":{"openalex":"https://openalex.org/W4402727885","doi":"https://doi.org/10.1109/cvpr52733.2024.02520"},"language":"en","primary_location":{"id":"doi:10.1109/cvpr52733.2024.02520","is_oa":false,"landing_page_url":"https://doi.org/10.1109/cvpr52733.2024.02520","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2024 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)","raw_type":"proceedings-article"},"type":"article","indexed_in":["crossref"],"open_access":{"is_oa":false,"oa_status":"closed","oa_url":null,"any_repository_has_fulltext":false},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5101701847","display_name":"Lin Ji","orcid":"https://orcid.org/0009-0000-8291-4022"},"institutions":[],"countries":[],"is_corresponding":true,"raw_author_name":"Ji Lin","raw_affiliation_strings":["NVIDIA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"NVIDIA","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5108868459","display_name":"Hongxu Yin","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Hongxu Yin","raw_affiliation_strings":["NVIDIA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"NVIDIA","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101947241","display_name":"Ping Wei","orcid":"https://orcid.org/0000-0002-8535-9527"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wei Ping","raw_affiliation_strings":["NVIDIA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"NVIDIA","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5066945976","display_name":"Pavlo Molchanov","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Pavlo Molchanov","raw_affiliation_strings":["NVIDIA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"NVIDIA","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5072436307","display_name":"Mohammad Shoeybi","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Mohammad Shoeybi","raw_affiliation_strings":["NVIDIA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"NVIDIA","institution_ids":[]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5086688453","display_name":"Song Han","orcid":"https://orcid.org/0000-0001-7758-3679"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Song Han","raw_affiliation_strings":["NVIDIA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"NVIDIA","institution_ids":[]}]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":6,"corresponding_author_ids":["https://openalex.org/A5101701847"],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":37.0613,"has_fulltext":false,"cited_by_count":161,"citation_normalized_percentile":{"value":0.99898759,"is_in_top_1_percent":true,"is_in_top_10_percent":true},"cited_by_percentile_year":{"min":99,"max":100},"biblio":{"volume":null,"issue":null,"first_page":"26679","last_page":"26689"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T11714","display_name":"Multimodal Machine Learning Applications","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/T11714","display_name":"Multimodal Machine Learning Applications","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/T10812","display_name":"Human Pose and Action Recognition","score":0.9972000122070312,"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/T10627","display_name":"Advanced Image and Video Retrieval Techniques","score":0.991100013256073,"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/training","display_name":"Training (meteorology)","score":0.6756824254989624},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.661537766456604},{"id":"https://openalex.org/keywords/natural-language-processing","display_name":"Natural language processing","score":0.4261581003665924},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.400075763463974},{"id":"https://openalex.org/keywords/geography","display_name":"Geography","score":0.0780867338180542}],"concepts":[{"id":"https://openalex.org/C2777211547","wikidata":"https://www.wikidata.org/wiki/Q17141490","display_name":"Training (meteorology)","level":2,"score":0.6756824254989624},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.661537766456604},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.4261581003665924},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.400075763463974},{"id":"https://openalex.org/C205649164","wikidata":"https://www.wikidata.org/wiki/Q1071","display_name":"Geography","level":0,"score":0.0780867338180542},{"id":"https://openalex.org/C153294291","wikidata":"https://www.wikidata.org/wiki/Q25261","display_name":"Meteorology","level":1,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/cvpr52733.2024.02520","is_oa":false,"landing_page_url":"https://doi.org/10.1109/cvpr52733.2024.02520","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2024 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)","raw_type":"proceedings-article"}],"best_oa_location":null,"sustainable_development_goals":[{"score":0.75,"id":"https://metadata.un.org/sdg/4","display_name":"Quality Education"}],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":96,"referenced_works":["https://openalex.org/W1861492603","https://openalex.org/W2185175083","https://openalex.org/W2560730294","https://openalex.org/W2947312908","https://openalex.org/W2963518342","https://openalex.org/W2963622213","https://openalex.org/W2964110616","https://openalex.org/W2979382951","https://openalex.org/W2990138404","https://openalex.org/W3121904249","https://openalex.org/W4221167110","https://openalex.org/W4224308101","https://openalex.org/W4225323055","https://openalex.org/W4225591000","https://openalex.org/W4226278401","https://openalex.org/W4229005866","https://openalex.org/W4286987939","https://openalex.org/W4287391717","https://openalex.org/W4288089799","https://openalex.org/W4296605665","https://openalex.org/W4306820534","https://openalex.org/W4307079201","https://openalex.org/W4318718936","https://openalex.org/W4322718191","https://openalex.org/W4323572061","https://openalex.org/W4361229539","https://openalex.org/W4366197770","https://openalex.org/W4366850747","https://openalex.org/W4367367040","https://openalex.org/W4376312115","https://openalex.org/W4378510493","https://openalex.org/W4378942562","https://openalex.org/W4380353763","https://openalex.org/W4382142077","https://openalex.org/W4382490555","https://openalex.org/W4382491206","https://openalex.org/W4384268538","https://openalex.org/W4384918448","https://openalex.org/W4385245566","https://openalex.org/W4385473806","https://openalex.org/W4385571157","https://openalex.org/W4385645323","https://openalex.org/W4386794522","https://openalex.org/W4387209488","https://openalex.org/W4387294588","https://openalex.org/W4388482029","https://openalex.org/W4388962905","https://openalex.org/W4389523832","https://openalex.org/W4395091069","https://openalex.org/W4402670859","https://openalex.org/W4402671548","https://openalex.org/W4402727764","https://openalex.org/W6764256271","https://openalex.org/W6769627184","https://openalex.org/W6778883912","https://openalex.org/W6782879696","https://openalex.org/W6788811087","https://openalex.org/W6791353385","https://openalex.org/W6800875267","https://openalex.org/W6805239564","https://openalex.org/W6810081322","https://openalex.org/W6810220367","https://openalex.org/W6810296985","https://openalex.org/W6810334672","https://openalex.org/W6810738896","https://openalex.org/W6811013733","https://openalex.org/W6843405348","https://openalex.org/W6846007759","https://openalex.org/W6846474394","https://openalex.org/W6847076894","https://openalex.org/W6849177959","https://openalex.org/W6850503672","https://openalex.org/W6850625674","https://openalex.org/W6851938174","https://openalex.org/W6851948999","https://openalex.org/W6851950068","https://openalex.org/W6852060543","https://openalex.org/W6852874933","https://openalex.org/W6853116092","https://openalex.org/W6853330684","https://openalex.org/W6853515732","https://openalex.org/W6854222408","https://openalex.org/W6854262950","https://openalex.org/W6854308264","https://openalex.org/W6854347851","https://openalex.org/W6854510102","https://openalex.org/W6854866820","https://openalex.org/W6855297460","https://openalex.org/W6856032414","https://openalex.org/W6856535631","https://openalex.org/W6856659896","https://openalex.org/W6856800273","https://openalex.org/W6857162426","https://openalex.org/W6857614378","https://openalex.org/W6858268588","https://openalex.org/W6859336385"],"related_works":["https://openalex.org/W4391375266","https://openalex.org/W2748952813","https://openalex.org/W230091440","https://openalex.org/W2390279801","https://openalex.org/W2233261550","https://openalex.org/W2358668433","https://openalex.org/W4396701345","https://openalex.org/W2810751659","https://openalex.org/W258997015","https://openalex.org/W3204019825"],"abstract_inverted_index":{"Visual":[0,144],"language":[1,12,38],"models":[2],"(VLMs)":[3],"rapidly":[4],"progressed":[5],"with":[6,27],"the":[7,25,36,42,57,95,124,151],"recent":[8],"success":[9],"of":[10,35,126,170],"large":[11],"models.":[13],"There":[14],"have":[15],"been":[16],"growing":[17],"efforts":[18],"on":[19,49,186],"visual":[20,28,37],"instruction":[21,113,119],"tuning":[22],"to":[23,45,115],"extend":[24],"LLM":[26,65],"inputs,":[29],"but":[30,87,129],"lacks":[31],"an":[32,136],"in-depth":[33],"study":[34],"pre-training":[39,62,81,99,138,164],"process,":[40],"where":[41],"model":[43,146],"learns":[44],"perform":[46],"joint":[47],"modeling":[48],"both":[50],"modalities.":[51],"In":[52],"this":[53],"work,":[54],"we":[55,140],"examine":[56],"design":[58],"options":[59],"for":[60,189],"VLM":[61,67,132],"by":[63],"augmenting":[64],"towards":[66],"through":[68],"step-by-step":[69],"controllable":[70],"comparisons.":[71],"We":[72],"introduce":[73],"three":[74],"main":[75,157],"findings:":[76],"(1)":[77],"freezing":[78],"LLMs":[79],"during":[80,118],"can":[82],"achieve":[83],"decent":[84],"zero-shot":[85],"performance,":[86],"lack":[88],"in-context":[89,176],"learning":[90],"capability,":[91],"which":[92],"requires":[93],"unfreezing":[94],"LLM;":[96],"(2)":[97],"interleaved":[98],"data":[100,114,117],"is":[101,183],"beneficial":[102],"whereas":[103],"image-text":[104,116],"pairs":[105],"alone":[106],"are":[107],"not":[108,121],"optimal;":[109],"(3)":[110],"re-blending":[111],"text-only":[112,127],"fine-tuning":[120],"only":[122],"remedies":[123],"degradation":[125],"tasks,":[128],"also":[130,165,184],"boosts":[131],"task":[133],"accuracy.":[134],"With":[135],"enhanced":[137,175],"recipe":[139],"build":[141],"VILA,":[142,171],"a":[143],"Language":[145],"family":[147],"that":[148],"consistently":[149],"outperforms":[150],"state-of-the-art":[152],"models,":[153],"e.g.,":[154],"LLaVA-1.5,":[155],"across":[156],"benchmarks":[158],"without":[159],"bells":[160],"and":[161,178],"whistles.":[162],"Multi-modal":[163],"helps":[166],"unveil":[167],"appealing":[168],"properties":[169],"including":[172],"multi-image":[173],"reasoning,":[174],"learning,":[177],"better":[179],"world":[180],"knowledge.":[181],"VILA":[182],"deployable":[185],"Jetson":[187],"Orin":[188],"on-device":[190],"VLM.":[191]},"counts_by_year":[{"year":2026,"cited_by_count":16},{"year":2025,"cited_by_count":136},{"year":2024,"cited_by_count":9}],"updated_date":"2026-05-07T13:39:58.223016","created_date":"2025-10-10T00:00:00"}