Multi-modal data label model training method and device and electronic equipment
The invention provides a multi-modal data label model training method and device and electronic equipment, and the method comprises the steps: constructing a to-be-trained candidate multi-modal data label model, and obtaining a multi-modal training sample; performing feature extraction and resamplin...
Gespeichert in:
| Hauptverfasser: | , , , , , , , , |
|---|---|
| Format: | Patent |
| Sprache: | chi ; eng |
| Schlagworte: | |
| Online-Zugang: | Volltext bestellen |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| container_end_page | |
|---|---|
| container_issue | |
| container_start_page | |
| container_title | |
| container_volume | |
| creator | FANG JUN WANG LI LIU PENGZHANG ZHANG LEZHONG GAO TIANHAO BAO YONGJUN LIU CHAO LIU HANYU CHEN CHAOFAN |
| description | The invention provides a multi-modal data label model training method and device and electronic equipment, and the method comprises the steps: constructing a to-be-trained candidate multi-modal data label model, and obtaining a multi-modal training sample; performing feature extraction and resampling on the multi-modal training sample to obtain a multi-modal splicing feature of the multi-modal training sample; obtaining a candidate multi-modal fusion prompt vector and a candidate label prompt vector corresponding to the multi-modal training sample; and obtaining a training loss set of the candidate multi-modal data label model based on the candidate multi-modal fusion prompt vector, the candidate label prompt vector and the multi-modal splicing feature, so as to carry out model training on the candidate multi-modal data label model until the training is finished, and obtaining a trained target multi-modal data label model. The stability of the candidate multi-modal data label model is improved, the noise filt |
| format | Patent |
| fullrecord | <record><control><sourceid>epo_EVB</sourceid><recordid>TN_cdi_epo_espacenet_CN118152813A</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><sourcerecordid>CN118152813A</sourcerecordid><originalsourceid>FETCH-epo_espacenet_CN118152813A3</originalsourceid><addsrcrecordid>eNqNirsKwkAQRbexEPUfxg9IsQYhrQTFRrGwD-PuVRdmHyYTv98gfoDNPYfDnZvLaRQNVcyehTwrk_ANQlOYVnsOKaQHRegze-LkyeMdHL4KgdM-p-AIrzGUiKRLM7uzDFj9uDDrw_7aHiuU3GEo7JCgXXu2trHbTWPrXf3P5wNulTcW</addsrcrecordid><sourcetype>Open Access Repository</sourcetype><iscdi>true</iscdi><recordtype>patent</recordtype></control><display><type>patent</type><title>Multi-modal data label model training method and device and electronic equipment</title><source>esp@cenet</source><creator>FANG JUN ; WANG LI ; LIU PENGZHANG ; ZHANG LEZHONG ; GAO TIANHAO ; BAO YONGJUN ; LIU CHAO ; LIU HANYU ; CHEN CHAOFAN</creator><creatorcontrib>FANG JUN ; WANG LI ; LIU PENGZHANG ; ZHANG LEZHONG ; GAO TIANHAO ; BAO YONGJUN ; LIU CHAO ; LIU HANYU ; CHEN CHAOFAN</creatorcontrib><description>The invention provides a multi-modal data label model training method and device and electronic equipment, and the method comprises the steps: constructing a to-be-trained candidate multi-modal data label model, and obtaining a multi-modal training sample; performing feature extraction and resampling on the multi-modal training sample to obtain a multi-modal splicing feature of the multi-modal training sample; obtaining a candidate multi-modal fusion prompt vector and a candidate label prompt vector corresponding to the multi-modal training sample; and obtaining a training loss set of the candidate multi-modal data label model based on the candidate multi-modal fusion prompt vector, the candidate label prompt vector and the multi-modal splicing feature, so as to carry out model training on the candidate multi-modal data label model until the training is finished, and obtaining a trained target multi-modal data label model. The stability of the candidate multi-modal data label model is improved, the noise filt</description><language>chi ; eng</language><subject>CALCULATING ; COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS ; COMPUTING ; COUNTING ; ELECTRIC DIGITAL DATA PROCESSING ; PHYSICS</subject><creationdate>2024</creationdate><oa>free_for_read</oa><woscitedreferencessubscribed>false</woscitedreferencessubscribed></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><linktohtml>$$Uhttps://worldwide.espacenet.com/publicationDetails/biblio?FT=D&date=20240607&DB=EPODOC&CC=CN&NR=118152813A$$EHTML$$P50$$Gepo$$Hfree_for_read</linktohtml><link.rule.ids>230,308,777,882,25545,76296</link.rule.ids><linktorsrc>$$Uhttps://worldwide.espacenet.com/publicationDetails/biblio?FT=D&date=20240607&DB=EPODOC&CC=CN&NR=118152813A$$EView_record_in_European_Patent_Office$$FView_record_in_$$GEuropean_Patent_Office$$Hfree_for_read</linktorsrc></links><search><creatorcontrib>FANG JUN</creatorcontrib><creatorcontrib>WANG LI</creatorcontrib><creatorcontrib>LIU PENGZHANG</creatorcontrib><creatorcontrib>ZHANG LEZHONG</creatorcontrib><creatorcontrib>GAO TIANHAO</creatorcontrib><creatorcontrib>BAO YONGJUN</creatorcontrib><creatorcontrib>LIU CHAO</creatorcontrib><creatorcontrib>LIU HANYU</creatorcontrib><creatorcontrib>CHEN CHAOFAN</creatorcontrib><title>Multi-modal data label model training method and device and electronic equipment</title><description>The invention provides a multi-modal data label model training method and device and electronic equipment, and the method comprises the steps: constructing a to-be-trained candidate multi-modal data label model, and obtaining a multi-modal training sample; performing feature extraction and resampling on the multi-modal training sample to obtain a multi-modal splicing feature of the multi-modal training sample; obtaining a candidate multi-modal fusion prompt vector and a candidate label prompt vector corresponding to the multi-modal training sample; and obtaining a training loss set of the candidate multi-modal data label model based on the candidate multi-modal fusion prompt vector, the candidate label prompt vector and the multi-modal splicing feature, so as to carry out model training on the candidate multi-modal data label model until the training is finished, and obtaining a trained target multi-modal data label model. The stability of the candidate multi-modal data label model is improved, the noise filt</description><subject>CALCULATING</subject><subject>COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS</subject><subject>COMPUTING</subject><subject>COUNTING</subject><subject>ELECTRIC DIGITAL DATA PROCESSING</subject><subject>PHYSICS</subject><fulltext>true</fulltext><rsrctype>patent</rsrctype><creationdate>2024</creationdate><recordtype>patent</recordtype><sourceid>EVB</sourceid><recordid>eNqNirsKwkAQRbexEPUfxg9IsQYhrQTFRrGwD-PuVRdmHyYTv98gfoDNPYfDnZvLaRQNVcyehTwrk_ANQlOYVnsOKaQHRegze-LkyeMdHL4KgdM-p-AIrzGUiKRLM7uzDFj9uDDrw_7aHiuU3GEo7JCgXXu2trHbTWPrXf3P5wNulTcW</recordid><startdate>20240607</startdate><enddate>20240607</enddate><creator>FANG JUN</creator><creator>WANG LI</creator><creator>LIU PENGZHANG</creator><creator>ZHANG LEZHONG</creator><creator>GAO TIANHAO</creator><creator>BAO YONGJUN</creator><creator>LIU CHAO</creator><creator>LIU HANYU</creator><creator>CHEN CHAOFAN</creator><scope>EVB</scope></search><sort><creationdate>20240607</creationdate><title>Multi-modal data label model training method and device and electronic equipment</title><author>FANG JUN ; WANG LI ; LIU PENGZHANG ; ZHANG LEZHONG ; GAO TIANHAO ; BAO YONGJUN ; LIU CHAO ; LIU HANYU ; CHEN CHAOFAN</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-epo_espacenet_CN118152813A3</frbrgroupid><rsrctype>patents</rsrctype><prefilter>patents</prefilter><language>chi ; eng</language><creationdate>2024</creationdate><topic>CALCULATING</topic><topic>COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS</topic><topic>COMPUTING</topic><topic>COUNTING</topic><topic>ELECTRIC DIGITAL DATA PROCESSING</topic><topic>PHYSICS</topic><toplevel>online_resources</toplevel><creatorcontrib>FANG JUN</creatorcontrib><creatorcontrib>WANG LI</creatorcontrib><creatorcontrib>LIU PENGZHANG</creatorcontrib><creatorcontrib>ZHANG LEZHONG</creatorcontrib><creatorcontrib>GAO TIANHAO</creatorcontrib><creatorcontrib>BAO YONGJUN</creatorcontrib><creatorcontrib>LIU CHAO</creatorcontrib><creatorcontrib>LIU HANYU</creatorcontrib><creatorcontrib>CHEN CHAOFAN</creatorcontrib><collection>esp@cenet</collection></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext_linktorsrc</fulltext></delivery><addata><au>FANG JUN</au><au>WANG LI</au><au>LIU PENGZHANG</au><au>ZHANG LEZHONG</au><au>GAO TIANHAO</au><au>BAO YONGJUN</au><au>LIU CHAO</au><au>LIU HANYU</au><au>CHEN CHAOFAN</au><format>patent</format><genre>patent</genre><ristype>GEN</ristype><title>Multi-modal data label model training method and device and electronic equipment</title><date>2024-06-07</date><risdate>2024</risdate><abstract>The invention provides a multi-modal data label model training method and device and electronic equipment, and the method comprises the steps: constructing a to-be-trained candidate multi-modal data label model, and obtaining a multi-modal training sample; performing feature extraction and resampling on the multi-modal training sample to obtain a multi-modal splicing feature of the multi-modal training sample; obtaining a candidate multi-modal fusion prompt vector and a candidate label prompt vector corresponding to the multi-modal training sample; and obtaining a training loss set of the candidate multi-modal data label model based on the candidate multi-modal fusion prompt vector, the candidate label prompt vector and the multi-modal splicing feature, so as to carry out model training on the candidate multi-modal data label model until the training is finished, and obtaining a trained target multi-modal data label model. The stability of the candidate multi-modal data label model is improved, the noise filt</abstract><oa>free_for_read</oa></addata></record> |
| fulltext | fulltext_linktorsrc |
| identifier | |
| ispartof | |
| issn | |
| language | chi ; eng |
| recordid | cdi_epo_espacenet_CN118152813A |
| source | esp@cenet |
| subjects | CALCULATING COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS COMPUTING COUNTING ELECTRIC DIGITAL DATA PROCESSING PHYSICS |
| title | Multi-modal data label model training method and device and electronic equipment |
| url | https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2025-01-19T20%3A31%3A10IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-epo_EVB&rft_val_fmt=info:ofi/fmt:kev:mtx:patent&rft.genre=patent&rft.au=FANG%20JUN&rft.date=2024-06-07&rft_id=info:doi/&rft_dat=%3Cepo_EVB%3ECN118152813A%3C/epo_EVB%3E%3Curl%3E%3C/url%3E&disable_directlink=true&sfx.directlink=off&sfx.report_link=0&rft_id=info:oai/&rft_id=info:pmid/&rfr_iscdi=true |