{"id":"https://openalex.org/W7214928347","doi":"https://doi.org/10.48550/arxiv.2609.36893","title":"Momentum-Coupled Rubric Adaptation for Detailed Image Captioning","display_name":"Momentum-Coupled Rubric Adaptation for Detailed Image Captioning","publication_year":2026,"publication_date":"2026-09-29","ids":{"openalex":"https://openalex.org/W7214928347","doi":"https://doi.org/10.48550/arxiv.2609.36893"},"language":"en","primary_location":{"id":"pmh:oai:arXiv.org:2609.36893","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2609.36893","pdf_url":"https://arxiv.org/pdf/2609.36893","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/2609.36893","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5053236610","display_name":"\u5180\u8d1e\u6587","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I154099455","display_name":"Shandong University","ror":"https://ror.org/0207yh398","country_code":"CN","type":"education","lineage":["https://openalex.org/I154099455"]},{"id":"https://openalex.org/I4401726856","display_name":"Xiaohongshu (Little Red Book) (China)","ror":"https://ror.org/02apw2076","country_code":"CN","type":"company","lineage":["https://openalex.org/I4401726856"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Ji, Zhenwen","raw_affiliation_strings":["Xiaohongshu Inc","the Joint SDU-NTU Centre for Artificial Intelligence Research (C-FAIR) , Shandong University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Xiaohongshu Inc","institution_ids":["https://openalex.org/I4401726856"]},{"raw_affiliation_string":"the Joint SDU-NTU Centre for Artificial Intelligence Research (C-FAIR) , Shandong University","institution_ids":["https://openalex.org/I154099455"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5153638987","display_name":"Lei Jin","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I4401726856","display_name":"Xiaohongshu (Little Red Book) (China)","ror":"https://ror.org/02apw2076","country_code":"CN","type":"company","lineage":["https://openalex.org/I4401726856"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Jin, Lei","raw_affiliation_strings":["Xiaohongshu Inc"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Xiaohongshu Inc","institution_ids":["https://openalex.org/I4401726856"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5133664730","display_name":"Shanyong Wang","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I4401726856","display_name":"Xiaohongshu (Little Red Book) (China)","ror":"https://ror.org/02apw2076","country_code":"CN","type":"company","lineage":["https://openalex.org/I4401726856"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Wang, Shanyong","raw_affiliation_strings":["Xiaohongshu Inc"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Xiaohongshu Inc","institution_ids":["https://openalex.org/I4401726856"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5108818837","display_name":"Jiaming Lu","orcid":"https://orcid.org/0009-0003-4212-8803","observed_orcids":["https://orcid.org/0009-0003-4212-8803"]},"institutions":[{"id":"https://openalex.org/I4401726856","display_name":"Xiaohongshu (Little Red Book) (China)","ror":"https://ror.org/02apw2076","country_code":"CN","type":"company","lineage":["https://openalex.org/I4401726856"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Lu, Jiaming","raw_affiliation_strings":["Xiaohongshu Inc"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Xiaohongshu Inc","institution_ids":["https://openalex.org/I4401726856"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5153605754","display_name":"Chengqiang Lu","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I4401726856","display_name":"Xiaohongshu (Little Red Book) (China)","ror":"https://ror.org/02apw2076","country_code":"CN","type":"company","lineage":["https://openalex.org/I4401726856"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Lu, Chengqiang","raw_affiliation_strings":["Xiaohongshu Inc"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Xiaohongshu Inc","institution_ids":["https://openalex.org/I4401726856"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5153577234","display_name":"Yi Wu","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I4401726856","display_name":"Xiaohongshu (Little Red Book) (China)","ror":"https://ror.org/02apw2076","country_code":"CN","type":"company","lineage":["https://openalex.org/I4401726856"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Wu, Yi","raw_affiliation_strings":["Xiaohongshu Inc"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Xiaohongshu Inc","institution_ids":["https://openalex.org/I4401726856"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5153576928","display_name":"Yao Hu","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I4401726856","display_name":"Xiaohongshu (Little Red Book) (China)","ror":"https://ror.org/02apw2076","country_code":"CN","type":"company","lineage":["https://openalex.org/I4401726856"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Hu, Yao","raw_affiliation_strings":["Xiaohongshu Inc"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Xiaohongshu Inc","institution_ids":["https://openalex.org/I4401726856"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5153676294","display_name":"Lizhen Cui","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I154099455","display_name":"Shandong University","ror":"https://ror.org/0207yh398","country_code":"CN","type":"education","lineage":["https://openalex.org/I154099455"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Cui, Lizhen","raw_affiliation_strings":["the Joint SDU-NTU Centre for Artificial Intelligence Research (C-FAIR) , Shandong University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"the Joint SDU-NTU Centre for Artificial Intelligence Research (C-FAIR) , Shandong University","institution_ids":["https://openalex.org/I154099455"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5153681886","display_name":"Yanyu Xu","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I154099455","display_name":"Shandong University","ror":"https://ror.org/0207yh398","country_code":"CN","type":"education","lineage":["https://openalex.org/I154099455"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Xu, Yanyu","raw_affiliation_strings":["the Joint SDU-NTU Centre for Artificial Intelligence Research (C-FAIR) , Shandong University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"the Joint SDU-NTU Centre for Artificial Intelligence Research (C-FAIR) , Shandong University","institution_ids":["https://openalex.org/I154099455"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":2,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":true,"cited_by_count":2,"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/T11714","display_name":"Multimodal Machine Learning Applications","score":0.9994999766349792,"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":0.9994999766349792,"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/T10775","display_name":"Generative Adversarial Networks and Image Synthesis","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"}},{"id":"https://openalex.org/T10181","display_name":"Natural Language Processing Techniques","score":0.0,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"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/multi-task-fine-tuning","display_name":"multi-task fine-tuning","score":0.6779999732971191},{"id":"https://openalex.org/keywords/momentum-model","display_name":"momentum model","score":0.4399999976158142},{"id":"https://openalex.org/keywords/exponentially-weighted-moving-average","display_name":"exponential moving average","score":0.6549999713897705},{"id":"https://openalex.org/keywords/vision-language-models","display_name":"vision-language models","score":0.7120000123977661}],"concepts":[{"id":"https://openalex.org/C111640148","wikidata":"https://www.wikidata.org/wiki/Q847349","display_name":"Rubric","level":2,"score":0.9739000201225281},{"id":"https://openalex.org/C157657479","wikidata":"https://www.wikidata.org/wiki/Q2367247","display_name":"Closed captioning","level":3,"score":0.7809000015258789},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.696399986743927},{"id":"https://openalex.org/C2780992000","wikidata":"https://www.wikidata.org/wiki/Q17016113","display_name":"Generator (circuit theory)","level":3,"score":0.5321000218391418},{"id":"https://openalex.org/C184898388","wikidata":"https://www.wikidata.org/wiki/Q1435712","display_name":"Pairwise comparison","level":2,"score":0.46480000019073486},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.4535999894142151},{"id":"https://openalex.org/C164226766","wikidata":"https://www.wikidata.org/wiki/Q7293202","display_name":"Rank (graph theory)","level":2,"score":0.41029998660087585},{"id":"https://openalex.org/C139807058","wikidata":"https://www.wikidata.org/wiki/Q352374","display_name":"Adaptation (eye)","level":2,"score":0.3865000009536743},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.384799987077713},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.3720000088214874},{"id":"https://openalex.org/C2779530757","wikidata":"https://www.wikidata.org/wiki/Q1207505","display_name":"Quality (philosophy)","level":2,"score":0.326200008392334},{"id":"https://openalex.org/C165064840","wikidata":"https://www.wikidata.org/wiki/Q1321061","display_name":"Matching (statistics)","level":2,"score":0.3215999901294708},{"id":"https://openalex.org/C127705205","wikidata":"https://www.wikidata.org/wiki/Q5748245","display_name":"Heuristics","level":2,"score":0.3084000051021576},{"id":"https://openalex.org/C3770464","wikidata":"https://www.wikidata.org/wiki/Q775963","display_name":"Smoothing","level":2,"score":0.28690001368522644},{"id":"https://openalex.org/C23123220","wikidata":"https://www.wikidata.org/wiki/Q816826","display_name":"Information retrieval","level":1,"score":0.28600001335144043},{"id":"https://openalex.org/C115961682","wikidata":"https://www.wikidata.org/wiki/Q860623","display_name":"Image (mathematics)","level":2,"score":0.2833000123500824},{"id":"https://openalex.org/C86037889","wikidata":"https://www.wikidata.org/wiki/Q4330127","display_name":"Learning to rank","level":3,"score":0.2709999978542328},{"id":"https://openalex.org/C2776359362","wikidata":"https://www.wikidata.org/wiki/Q2145286","display_name":"Representation (politics)","level":3,"score":0.2529999911785126}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:arXiv.org:2609.36893","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2609.36893","pdf_url":"https://arxiv.org/pdf/2609.36893","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.2609.36893","doi":"https://doi.org/10.48550/arxiv.2609.36893","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2609.36893","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":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"pmh:oai:arXiv.org:2609.36893","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2609.36893","pdf_url":"https://arxiv.org/pdf/2609.36893","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/W7214928347.pdf","grobid_xml":"https://content.openalex.org/works/W7214928347.grobid-xml"},"referenced_works_count":38,"referenced_works":["https://openalex.org/W2101105183","https://openalex.org/W2892087592","https://openalex.org/W2908510526","https://openalex.org/W2950201573","https://openalex.org/W2952574180","https://openalex.org/W3120043490","https://openalex.org/W3153469116","https://openalex.org/W4226182655","https://openalex.org/W4285255856","https://openalex.org/W4389524413","https://openalex.org/W4402684027","https://openalex.org/W4403923795","https://openalex.org/W4413147844","https://openalex.org/W4414490998","https://openalex.org/W4414684918","https://openalex.org/W4414886170","https://openalex.org/W4415235458","https://openalex.org/W4415318414","https://openalex.org/W4415798334","https://openalex.org/W4416048728","https://openalex.org/W4416531666","https://openalex.org/W4416780398","https://openalex.org/W4417527299","https://openalex.org/W6891858849","https://openalex.org/W6891920207","https://openalex.org/W6948151711","https://openalex.org/W7083718151","https://openalex.org/W7106767948","https://openalex.org/W7127272498","https://openalex.org/W7129780791","https://openalex.org/W7131634831","https://openalex.org/W7134898772","https://openalex.org/W7154506115","https://openalex.org/W7160507623","https://openalex.org/W7161978236","https://openalex.org/W7162788548","https://openalex.org/W7164080106","https://openalex.org/W7170175357"],"related_works":["https://openalex.org/W7134898772","https://openalex.org/W7160847285","https://openalex.org/W7131634831","https://openalex.org/W7083704284","https://openalex.org/W2729842244","https://openalex.org/W4226257165","https://openalex.org/W4414566324","https://openalex.org/W2974212192","https://openalex.org/W7204587146","https://openalex.org/W7164080106"],"abstract_inverted_index":{"Detailed":[0],"image":[1,154],"captioning":[2,226],"requires":[3],"accurate":[4],"and":[5,20,44,61,81,97,135,153,161,195,245],"comprehensive":[6],"descriptions":[7],"of":[8,100,182,236],"fine-grained":[9],"visual":[10],"content,":[11],"yet":[12],"caption":[13,57,79],"quality":[14],"spans":[15],"factual":[16],"accuracy,":[17],"information":[18],"coverage,":[19],"clarity.":[21],"Compared":[22],"with":[23,102],"conventional":[24],"methods":[25,51,74],"that":[26,112,215],"rely":[27],"mainly":[28],"on":[29],"high-quality":[30],"supervision":[31],"or":[32],"holistic":[33],"rewards,":[34,160],"rubric-based":[35,159],"reinforcement":[36],"learning":[37],"decomposes":[38],"these":[39,114],"requirements":[40],"into":[41],"explicit":[42],"criteria":[43],"provides":[45,158],"targeted,":[46],"structured":[47],"feedback.":[48],"However,":[49],"existing":[50],"often":[52],"use":[53,177],"separate":[54,208],"models":[55],"for":[56],"generation,":[58],"rubric":[59,73,82,95,202,219],"construction,":[60],"judging,":[62],"which":[63],"may":[64,92],"lead":[65],"to":[66,127,186],"inconsistent":[67],"interpretations":[68],"across":[69],"roles.":[70,115],"Some":[71],"dynamic":[72],"alternate":[75],"updates":[76,170,206],"between":[77],"the":[78,86,130,139,148,163,172,183,193,238,246],"policy":[80,103],"generator":[83],"while":[84,211],"keeping":[85],"judge":[87],"fixed,":[88],"but":[89],"staged":[90],"optimization":[91,210],"still":[93],"leave":[94],"construction":[96],"judging":[98],"out":[99],"step":[101],"optimization.":[104],"We":[105],"propose":[106],"MoCo":[107,228],"Rubric,":[108],"a":[109,123],"two-stage":[110],"framework":[111],"coordinates":[113],"First,":[116],"role-conditioned,":[117],"shared-parameter":[118],"multi-task":[119],"supervised":[120],"fine-tuning":[121],"equips":[122],"single":[124],"vision--language":[125],"model":[126,190],"serve":[128],"as":[129],"Caption":[131],"Policy,":[132,150],"Rubric":[133,136,229],"Generator,":[134],"Judge.":[137,196],"Then,":[138],"Generator":[140,194],"constructs":[141],"rubrics":[142],"online":[143],"from":[144],"captions":[145],"sampled":[146],"by":[147,192],"current":[149],"reference":[151],"captions,":[152],"evidence.":[155],"The":[156],"Judge":[157],"only":[162],"Policy":[164,169,184,205],"receives":[165],"GRPO":[166],"updates.":[167],"As":[168],"change":[171],"candidates":[173],"being":[174],"evaluated,":[175],"we":[176],"an":[178,231],"exponential":[179],"moving":[180],"average":[181,232,248],"parameters":[185],"update":[187],"one":[188],"momentum":[189],"shared":[191],"This":[197],"gradual":[198],"transfer":[199],"lets":[200],"both":[201],"roles":[203],"track":[204],"without":[207],"RL":[209],"smoothing":[212],"parameter":[213],"changes":[214],"could":[216],"disrupt":[217],"their":[218],"capabilities":[220],"under":[221],"direct":[222],"synchronization.":[223],"Across":[224],"five":[225],"benchmarks,":[227],"achieves":[230],"pairwise":[233],"win":[234],"rate":[235],"72.83\\%,":[237],"best":[239],"mean":[240],"rank":[241],"in":[242,250],"blind":[243],"ranking,":[244],"highest":[247],"score":[249],"caption-based":[251],"question":[252],"answering.":[253]},"counts_by_year":[{"year":2026,"cited_by_count":2}],"updated_date":"2026-11-07T11:25:07.637180","created_date":"2026-10-01T00:00:00"}