TRAINING OF TEXT AND IMAGE MODELS
A method of training a text model using a plurality of text passage combinations, each text passage combination comprising a respective first text passage and a respective second text passage describing a same matter as the respective first text passage but being differently worded than the respecti...
Gespeichert in:
| Hauptverfasser: | , , , , , , , , , |
|---|---|
| Format: | Patent |
| Sprache: | eng ; fre ; ger |
| 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 | NAUMANN, Tristan SCHWAIGHOFER, Anton POON, Hoifung BANNUR, Shruthi Jaisimha HYLAND, Stephanie OKTAY, Ozan COELHO DE CASTRO, Daniel VALLE, Javier Alvarez USUYAMA, Naoto NORI, Aditya |
| description | A method of training a text model using a plurality of text passage combinations, each text passage combination comprising a respective first text passage and a respective second text passage describing a same matter as the respective first text passage but being differently worded than the respective first text passage. The training comprises minimizing a measure of statistical difference between a respective value of a first text embedding and the corresponding value of a second text embedding over the plurality of text passage combinations. The method then comprises jointly training the text model and an image model based on plurality of image-text combinations, each comprising a respective image and a respective textual report describing the respective image. The joint training comprises minimizing a measure of statistical difference between the value of an image embedding and the corresponding value of a third text embedding over the plurality of image-text combinations. |
| format | Patent |
| fullrecord | <record><control><sourceid>epo_EVB</sourceid><recordid>TN_cdi_epo_espacenet_EP4266195A1</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><sourcerecordid>EP4266195A1</sourcerecordid><originalsourceid>FETCH-epo_espacenet_EP4266195A13</originalsourceid><addsrcrecordid>eNrjZFAMCXL09PP0c1fwd1MIcY0IUXD0c1Hw9HV0d1Xw9Xdx9QnmYWBNS8wpTuWF0twMCm6uIc4euqkF-fGpxQWJyal5qSXxrgEmRmZmhpamjobGRCgBALfmIYI</addsrcrecordid><sourcetype>Open Access Repository</sourcetype><iscdi>true</iscdi><recordtype>patent</recordtype></control><display><type>patent</type><title>TRAINING OF TEXT AND IMAGE MODELS</title><source>esp@cenet</source><creator>NAUMANN, Tristan ; SCHWAIGHOFER, Anton ; POON, Hoifung ; BANNUR, Shruthi Jaisimha ; HYLAND, Stephanie ; OKTAY, Ozan ; COELHO DE CASTRO, Daniel ; VALLE, Javier Alvarez ; USUYAMA, Naoto ; NORI, Aditya</creator><creatorcontrib>NAUMANN, Tristan ; SCHWAIGHOFER, Anton ; POON, Hoifung ; BANNUR, Shruthi Jaisimha ; HYLAND, Stephanie ; OKTAY, Ozan ; COELHO DE CASTRO, Daniel ; VALLE, Javier Alvarez ; USUYAMA, Naoto ; NORI, Aditya</creatorcontrib><description>A method of training a text model using a plurality of text passage combinations, each text passage combination comprising a respective first text passage and a respective second text passage describing a same matter as the respective first text passage but being differently worded than the respective first text passage. The training comprises minimizing a measure of statistical difference between a respective value of a first text embedding and the corresponding value of a second text embedding over the plurality of text passage combinations. The method then comprises jointly training the text model and an image model based on plurality of image-text combinations, each comprising a respective image and a respective textual report describing the respective image. The joint training comprises minimizing a measure of statistical difference between the value of an image embedding and the corresponding value of a third text embedding over the plurality of image-text combinations.</description><language>eng ; fre ; ger</language><subject>CALCULATING ; COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS ; COMPUTING ; COUNTING ; ELECTRIC DIGITAL DATA PROCESSING ; HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATIONTECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING ORPROCESSING OF MEDICAL OR HEALTHCARE DATA ; INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTEDFOR SPECIFIC APPLICATION FIELDS ; PHYSICS</subject><creationdate>2023</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=20231025&DB=EPODOC&CC=EP&NR=4266195A1$$EHTML$$P50$$Gepo$$Hfree_for_read</linktohtml><link.rule.ids>230,308,780,885,25564,76547</link.rule.ids><linktorsrc>$$Uhttps://worldwide.espacenet.com/publicationDetails/biblio?FT=D&date=20231025&DB=EPODOC&CC=EP&NR=4266195A1$$EView_record_in_European_Patent_Office$$FView_record_in_$$GEuropean_Patent_Office$$Hfree_for_read</linktorsrc></links><search><creatorcontrib>NAUMANN, Tristan</creatorcontrib><creatorcontrib>SCHWAIGHOFER, Anton</creatorcontrib><creatorcontrib>POON, Hoifung</creatorcontrib><creatorcontrib>BANNUR, Shruthi Jaisimha</creatorcontrib><creatorcontrib>HYLAND, Stephanie</creatorcontrib><creatorcontrib>OKTAY, Ozan</creatorcontrib><creatorcontrib>COELHO DE CASTRO, Daniel</creatorcontrib><creatorcontrib>VALLE, Javier Alvarez</creatorcontrib><creatorcontrib>USUYAMA, Naoto</creatorcontrib><creatorcontrib>NORI, Aditya</creatorcontrib><title>TRAINING OF TEXT AND IMAGE MODELS</title><description>A method of training a text model using a plurality of text passage combinations, each text passage combination comprising a respective first text passage and a respective second text passage describing a same matter as the respective first text passage but being differently worded than the respective first text passage. The training comprises minimizing a measure of statistical difference between a respective value of a first text embedding and the corresponding value of a second text embedding over the plurality of text passage combinations. The method then comprises jointly training the text model and an image model based on plurality of image-text combinations, each comprising a respective image and a respective textual report describing the respective image. The joint training comprises minimizing a measure of statistical difference between the value of an image embedding and the corresponding value of a third text embedding over the plurality of image-text combinations.</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>HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATIONTECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING ORPROCESSING OF MEDICAL OR HEALTHCARE DATA</subject><subject>INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTEDFOR SPECIFIC APPLICATION FIELDS</subject><subject>PHYSICS</subject><fulltext>true</fulltext><rsrctype>patent</rsrctype><creationdate>2023</creationdate><recordtype>patent</recordtype><sourceid>EVB</sourceid><recordid>eNrjZFAMCXL09PP0c1fwd1MIcY0IUXD0c1Hw9HV0d1Xw9Xdx9QnmYWBNS8wpTuWF0twMCm6uIc4euqkF-fGpxQWJyal5qSXxrgEmRmZmhpamjobGRCgBALfmIYI</recordid><startdate>20231025</startdate><enddate>20231025</enddate><creator>NAUMANN, Tristan</creator><creator>SCHWAIGHOFER, Anton</creator><creator>POON, Hoifung</creator><creator>BANNUR, Shruthi Jaisimha</creator><creator>HYLAND, Stephanie</creator><creator>OKTAY, Ozan</creator><creator>COELHO DE CASTRO, Daniel</creator><creator>VALLE, Javier Alvarez</creator><creator>USUYAMA, Naoto</creator><creator>NORI, Aditya</creator><scope>EVB</scope></search><sort><creationdate>20231025</creationdate><title>TRAINING OF TEXT AND IMAGE MODELS</title><author>NAUMANN, Tristan ; SCHWAIGHOFER, Anton ; POON, Hoifung ; BANNUR, Shruthi Jaisimha ; HYLAND, Stephanie ; OKTAY, Ozan ; COELHO DE CASTRO, Daniel ; VALLE, Javier Alvarez ; USUYAMA, Naoto ; NORI, Aditya</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-epo_espacenet_EP4266195A13</frbrgroupid><rsrctype>patents</rsrctype><prefilter>patents</prefilter><language>eng ; fre ; ger</language><creationdate>2023</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>HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATIONTECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING ORPROCESSING OF MEDICAL OR HEALTHCARE DATA</topic><topic>INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTEDFOR SPECIFIC APPLICATION FIELDS</topic><topic>PHYSICS</topic><toplevel>online_resources</toplevel><creatorcontrib>NAUMANN, Tristan</creatorcontrib><creatorcontrib>SCHWAIGHOFER, Anton</creatorcontrib><creatorcontrib>POON, Hoifung</creatorcontrib><creatorcontrib>BANNUR, Shruthi Jaisimha</creatorcontrib><creatorcontrib>HYLAND, Stephanie</creatorcontrib><creatorcontrib>OKTAY, Ozan</creatorcontrib><creatorcontrib>COELHO DE CASTRO, Daniel</creatorcontrib><creatorcontrib>VALLE, Javier Alvarez</creatorcontrib><creatorcontrib>USUYAMA, Naoto</creatorcontrib><creatorcontrib>NORI, Aditya</creatorcontrib><collection>esp@cenet</collection></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext_linktorsrc</fulltext></delivery><addata><au>NAUMANN, Tristan</au><au>SCHWAIGHOFER, Anton</au><au>POON, Hoifung</au><au>BANNUR, Shruthi Jaisimha</au><au>HYLAND, Stephanie</au><au>OKTAY, Ozan</au><au>COELHO DE CASTRO, Daniel</au><au>VALLE, Javier Alvarez</au><au>USUYAMA, Naoto</au><au>NORI, Aditya</au><format>patent</format><genre>patent</genre><ristype>GEN</ristype><title>TRAINING OF TEXT AND IMAGE MODELS</title><date>2023-10-25</date><risdate>2023</risdate><abstract>A method of training a text model using a plurality of text passage combinations, each text passage combination comprising a respective first text passage and a respective second text passage describing a same matter as the respective first text passage but being differently worded than the respective first text passage. The training comprises minimizing a measure of statistical difference between a respective value of a first text embedding and the corresponding value of a second text embedding over the plurality of text passage combinations. The method then comprises jointly training the text model and an image model based on plurality of image-text combinations, each comprising a respective image and a respective textual report describing the respective image. The joint training comprises minimizing a measure of statistical difference between the value of an image embedding and the corresponding value of a third text embedding over the plurality of image-text combinations.</abstract><oa>free_for_read</oa></addata></record> |
| fulltext | fulltext_linktorsrc |
| identifier | |
| ispartof | |
| issn | |
| language | eng ; fre ; ger |
| recordid | cdi_epo_espacenet_EP4266195A1 |
| source | esp@cenet |
| subjects | CALCULATING COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS COMPUTING COUNTING ELECTRIC DIGITAL DATA PROCESSING HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATIONTECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING ORPROCESSING OF MEDICAL OR HEALTHCARE DATA INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTEDFOR SPECIFIC APPLICATION FIELDS PHYSICS |
| title | TRAINING OF TEXT AND IMAGE MODELS |
| url | https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2025-01-04T17%3A43%3A14IST&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=NAUMANN,%20Tristan&rft.date=2023-10-25&rft_id=info:doi/&rft_dat=%3Cepo_EVB%3EEP4266195A1%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 |