{"id":"https://openalex.org/W4403780665","doi":"https://doi.org/10.1145/3664647.3680686","title":"MoS<sup>2</sup>: Mixture of Scale and Shift Experts for Text-Only Video Captioning","display_name":"MoS<sup>2</sup>: Mixture of Scale and Shift Experts for Text-Only Video Captioning","publication_year":2024,"publication_date":"2024-10-26","ids":{"openalex":"https://openalex.org/W4403780665","doi":"https://doi.org/10.1145/3664647.3680686"},"language":"en","primary_location":{"id":"doi:10.1145/3664647.3680686","is_oa":false,"landing_page_url":"https://doi.org/10.1145/3664647.3680686","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 32nd ACM International Conference on Multimedia","raw_type":"proceedings-article"},"type":"conference-paper","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/A5048926828","display_name":"Heng Jia","orcid":"https://orcid.org/0000-0001-8062-9455"},"institutions":[{"id":"https://openalex.org/I76130692","display_name":"Zhejiang University","ror":"https://ror.org/00a2xv884","country_code":"CN","type":"education","lineage":["https://openalex.org/I76130692"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Heng Jia","raw_affiliation_strings":["ReLER, CCAI, Zhejiang University, Hangzhou, China"],"raw_orcid":"https://orcid.org/0000-0001-8062-9455","affiliations":[{"raw_affiliation_string":"ReLER, CCAI, Zhejiang University, Hangzhou, China","institution_ids":["https://openalex.org/I76130692"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101789941","display_name":"Yunqiu Xu","orcid":"https://orcid.org/0000-0002-2940-4805"},"institutions":[{"id":"https://openalex.org/I76130692","display_name":"Zhejiang University","ror":"https://ror.org/00a2xv884","country_code":"CN","type":"education","lineage":["https://openalex.org/I76130692"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Yunqiu Xu","raw_affiliation_strings":["ReLER, CCAI, Zhejiang University, Hangzhou, China"],"raw_orcid":"https://orcid.org/0000-0002-2940-4805","affiliations":[{"raw_affiliation_string":"ReLER, CCAI, Zhejiang University, Hangzhou, China","institution_ids":["https://openalex.org/I76130692"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5043617790","display_name":"Linchao Zhu","orcid":"https://orcid.org/0000-0002-4093-7557"},"institutions":[{"id":"https://openalex.org/I76130692","display_name":"Zhejiang University","ror":"https://ror.org/00a2xv884","country_code":"CN","type":"education","lineage":["https://openalex.org/I76130692"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Linchao Zhu","raw_affiliation_strings":["The State Key Lab of Brain-Machine Intelligence, Zhejiang University, Hangzhou, China"],"raw_orcid":"https://orcid.org/0000-0002-4093-7557","affiliations":[{"raw_affiliation_string":"The State Key Lab of Brain-Machine Intelligence, Zhejiang University, Hangzhou, China","institution_ids":["https://openalex.org/I76130692"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5102992722","display_name":"Guang Chen","orcid":"https://orcid.org/0000-0003-1487-8044"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Guang Chen","raw_affiliation_strings":["ByteDance Inc., San Jose, CA, USA"],"raw_orcid":"https://orcid.org/0000-0003-1487-8044","affiliations":[{"raw_affiliation_string":"ByteDance Inc., San Jose, CA, USA","institution_ids":[]}]},{"author_position":"middle","author":{"id":null,"display_name":"Yufei Wang","orcid":"https://orcid.org/0009-0008-4407-4948"},"institutions":[{"id":"https://openalex.org/I4210099336","display_name":"Menlo School","ror":"https://ror.org/01240pn49","country_code":"US","type":"education","lineage":["https://openalex.org/I4210099336"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Yufei Wang","raw_affiliation_strings":["character.ai, Menlo Park, CA, USA"],"raw_orcid":"https://orcid.org/0009-0008-4407-4948","affiliations":[{"raw_affiliation_string":"character.ai, Menlo Park, CA, USA","institution_ids":["https://openalex.org/I4210099336"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5005421447","display_name":"Yi Yang","orcid":"https://orcid.org/0000-0002-0512-880X"},"institutions":[{"id":"https://openalex.org/I76130692","display_name":"Zhejiang University","ror":"https://ror.org/00a2xv884","country_code":"CN","type":"education","lineage":["https://openalex.org/I76130692"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Yi Yang","raw_affiliation_strings":["ReLER, CCAI, Zhejiang University, Hangzhou, China"],"raw_orcid":"https://orcid.org/0000-0002-0512-880X","affiliations":[{"raw_affiliation_string":"ReLER, CCAI, Zhejiang University, Hangzhou, China","institution_ids":["https://openalex.org/I76130692"]}]}],"institutions":[],"countries_distinct_count":2,"institutions_distinct_count":2,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":1.5776,"has_fulltext":false,"cited_by_count":6,"citation_normalized_percentile":{"value":0.8817263,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":{"min":90,"max":98},"biblio":{"volume":null,"issue":null,"first_page":"8498","last_page":"8507"},"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/T11439","display_name":"Video Analysis and Summarization","score":0.996399998664856,"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":0.9944999814033508,"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/closed-captioning","display_name":"Closed captioning","score":0.9506179094314575},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.6737476587295532},{"id":"https://openalex.org/keywords/scale","display_name":"Scale (ratio)","score":0.580696702003479},{"id":"https://openalex.org/keywords/natural-language-processing","display_name":"Natural language processing","score":0.41225215792655945},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.36708372831344604},{"id":"https://openalex.org/keywords/speech-recognition","display_name":"Speech recognition","score":0.35135865211486816},{"id":"https://openalex.org/keywords/multimedia","display_name":"Multimedia","score":0.3427277207374573},{"id":"https://openalex.org/keywords/image","display_name":"Image (mathematics)","score":0.13512516021728516},{"id":"https://openalex.org/keywords/physics","display_name":"Physics","score":0.08140614628791809}],"concepts":[{"id":"https://openalex.org/C157657479","wikidata":"https://www.wikidata.org/wiki/Q2367247","display_name":"Closed captioning","level":3,"score":0.9506179094314575},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6737476587295532},{"id":"https://openalex.org/C2778755073","wikidata":"https://www.wikidata.org/wiki/Q10858537","display_name":"Scale (ratio)","level":2,"score":0.580696702003479},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.41225215792655945},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.36708372831344604},{"id":"https://openalex.org/C28490314","wikidata":"https://www.wikidata.org/wiki/Q189436","display_name":"Speech recognition","level":1,"score":0.35135865211486816},{"id":"https://openalex.org/C49774154","wikidata":"https://www.wikidata.org/wiki/Q131765","display_name":"Multimedia","level":1,"score":0.3427277207374573},{"id":"https://openalex.org/C115961682","wikidata":"https://www.wikidata.org/wiki/Q860623","display_name":"Image (mathematics)","level":2,"score":0.13512516021728516},{"id":"https://openalex.org/C121332964","wikidata":"https://www.wikidata.org/wiki/Q413","display_name":"Physics","level":0,"score":0.08140614628791809},{"id":"https://openalex.org/C62520636","wikidata":"https://www.wikidata.org/wiki/Q944","display_name":"Quantum mechanics","level":1,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1145/3664647.3680686","is_oa":false,"landing_page_url":"https://doi.org/10.1145/3664647.3680686","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 32nd ACM International Conference on Multimedia","raw_type":"proceedings-article"}],"best_oa_location":null,"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":27,"referenced_works":["https://openalex.org/W1593271688","https://openalex.org/W1956340063","https://openalex.org/W2108325777","https://openalex.org/W2251935656","https://openalex.org/W2425121537","https://openalex.org/W2600463316","https://openalex.org/W2914695977","https://openalex.org/W2989322838","https://openalex.org/W3132057951","https://openalex.org/W3217340782","https://openalex.org/W4212774754","https://openalex.org/W4225323055","https://openalex.org/W4306294758","https://openalex.org/W4312463400","https://openalex.org/W4312746376","https://openalex.org/W4312884055","https://openalex.org/W4312924260","https://openalex.org/W4312938727","https://openalex.org/W4319049530","https://openalex.org/W4386047824","https://openalex.org/W4391941517","https://openalex.org/W4393147969","https://openalex.org/W4393159300","https://openalex.org/W4402716170","https://openalex.org/W4403791865","https://openalex.org/W4404724811","https://openalex.org/W6767976850"],"related_works":["https://openalex.org/W4210416330","https://openalex.org/W2775506363","https://openalex.org/W3088136942","https://openalex.org/W4290852288","https://openalex.org/W2949362007","https://openalex.org/W4283207562","https://openalex.org/W2963177403","https://openalex.org/W2330246314","https://openalex.org/W2949522393","https://openalex.org/W4289422896"],"abstract_inverted_index":{"Video":[0],"captioning":[1,38,121],"is":[2,22],"a":[3,32,70,83,133],"challenging":[4],"task":[5],"and":[6,24,58,88,111,137,159,166,197,200],"typically":[7],"requires":[8],"paired":[9,205],"video-text":[10],"data":[11,157,206],"for":[12,20,119],"training.":[13],"However,":[14],"manually":[15],"annotating":[16],"coherent":[17],"textual":[18],"descriptions":[19],"videos":[21],"laborious":[23],"time-consuming.":[25],"To":[26,101],"address":[27],"this":[28],"challenge,":[29],"we":[30,55,106],"propose":[31],"novel":[33],"approach":[34,68],"that":[35],"enhances":[36,151],"video":[37,60],"using":[39],"only":[40,188],"synthetic":[41,99,165,189],"text":[42,47,190],"data.":[43,100],"Leveraging":[44],"the":[45,64,93,144,152,161,173,193,202,208],"exceptional":[46],"generation":[48],"capabilities":[49],"of":[50,85,109,135,175],"large":[51],"language":[52],"models":[53,199],"(LLMs),":[54],"produce":[56],"high-quality":[57,86],"diverse":[59],"captions":[61,80,95],"tailored":[62],"to":[63,81,96,128,143,155],"target":[65,209],"domain.":[66,210],"Our":[67,182],"employs":[69,124],"two-stage":[71],"prompting":[72,91],"strategy:":[73],"first":[74],"prompt":[75],"GPT-4":[76],"with":[77,92,187],"few-shot":[78],"target-domain":[79],"create":[82],"set":[84],"captions,":[87,105],"then":[89],"continue":[90],"generated":[94],"acquire":[97],"large-scale":[98],"effectively":[102],"utilize":[103],"these":[104],"introduce":[107],"Mixture":[108],"Scale":[110],"Shift":[112],"experts":[113],"(MoS2),":[114],"an":[115],"efficient":[116,180],"adaptation":[117],"method":[118,171,183],"pre-trained":[120],"models.":[122],"MoS2":[123],"lightweight":[125],"routing":[126],"networks":[127],"estimate":[129],"probability":[130],"distributions":[131],"over":[132],"collection":[134],"scale":[136],"shift":[138,163],"experts,":[139],"dynamically":[140],"allocating":[141],"tokens":[142],"appropriate":[145],"experts.":[146],"This":[147],"dynamic":[148],"adjustment":[149],"mechanism":[150],"model's":[153],"ability":[154],"handle":[156],"variations":[158],"mitigates":[160],"distribution":[162],"between":[164,195],"real":[167],"captions.":[168],"Moreover,":[169],"our":[170],"reduces":[172],"number":[174],"learnable":[176],"parameters,":[177],"facilitating":[178],"more":[179],"adaptation.":[181],"achieves":[184],"superior":[185],"performance":[186],"data,":[191],"narrowing":[192],"gap":[194],"zero-shot":[196],"fine-tuned":[198],"reducing":[201],"dependency":[203],"on":[204],"from":[207]},"counts_by_year":[{"year":2026,"cited_by_count":2},{"year":2025,"cited_by_count":3},{"year":2024,"cited_by_count":1}],"updated_date":"2026-07-29T14:22:42.915294","created_date":"2025-10-10T00:00:00"}