Neural Language Modeling with Visual Features
Multimodal language models attempt to incorporate non-linguistic features for the language modeling task. In this work, we extend a standard recurrent neural network (RNN) language model with features derived from videos. We train our models on data that is two orders-of-magnitude bigger than datase...
Gespeichert in:
Hauptverfasser: | , , |
---|---|
Format: | Artikel |
Sprache: | 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 | Anastasopoulos, Antonios Kumar, Shankar Liao, Hank |
description | Multimodal language models attempt to incorporate non-linguistic features for
the language modeling task. In this work, we extend a standard recurrent neural
network (RNN) language model with features derived from videos. We train our
models on data that is two orders-of-magnitude bigger than datasets used in
prior work. We perform a thorough exploration of model architectures for
combining visual and text features. Our experiments on two corpora (YouCookII
and 20bn-something-something-v2) show that the best performing architecture
consists of middle fusion of visual and text features, yielding over 25%
relative improvement in perplexity. We report analysis that provides insights
into why our multimodal language model improves upon a standard RNN language
model. |
doi_str_mv | 10.48550/arxiv.1903.02930 |
format | Article |
fullrecord | <record><control><sourceid>arxiv_GOX</sourceid><recordid>TN_cdi_arxiv_primary_1903_02930</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><sourcerecordid>1903_02930</sourcerecordid><originalsourceid>FETCH-LOGICAL-a670-80cf513c2f820a289c61a946ef5adc6709bccbcff07cfabffe96cfceb5224c253</originalsourceid><addsrcrecordid>eNotzrFuwkAQBNBrKBDkA6jiH7CzvvPZvhIhCJFM0qC01nq965zkADrjJPx9CKGaYkajp9QihSQrrYUnDD_-K0kdmAS0MzBV8SuPAfuowkM3YsfR7thy7w9d9O3PH9G7H8Zru2E8j4GHuZoI9gM_3HOm9pv1frWNq7fnl9WyijEvIC6BxKaGtJQaUJeO8hRdlrNYbOm6cA1RQyJQkGAjwi4nIW6s1hlpa2bq8f_25q1PwX9iuNR_7vrmNr90eT3y</addsrcrecordid><sourcetype>Open Access Repository</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype></control><display><type>article</type><title>Neural Language Modeling with Visual Features</title><source>arXiv.org</source><creator>Anastasopoulos, Antonios ; Kumar, Shankar ; Liao, Hank</creator><creatorcontrib>Anastasopoulos, Antonios ; Kumar, Shankar ; Liao, Hank</creatorcontrib><description>Multimodal language models attempt to incorporate non-linguistic features for
the language modeling task. In this work, we extend a standard recurrent neural
network (RNN) language model with features derived from videos. We train our
models on data that is two orders-of-magnitude bigger than datasets used in
prior work. We perform a thorough exploration of model architectures for
combining visual and text features. Our experiments on two corpora (YouCookII
and 20bn-something-something-v2) show that the best performing architecture
consists of middle fusion of visual and text features, yielding over 25%
relative improvement in perplexity. We report analysis that provides insights
into why our multimodal language model improves upon a standard RNN language
model.</description><identifier>DOI: 10.48550/arxiv.1903.02930</identifier><language>eng</language><subject>Computer Science - Computation and Language</subject><creationdate>2019-03</creationdate><rights>http://arxiv.org/licenses/nonexclusive-distrib/1.0</rights><oa>free_for_read</oa><woscitedreferencessubscribed>false</woscitedreferencessubscribed></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><link.rule.ids>228,230,780,885</link.rule.ids><linktorsrc>$$Uhttps://arxiv.org/abs/1903.02930$$EView_record_in_Cornell_University$$FView_record_in_$$GCornell_University$$Hfree_for_read</linktorsrc><backlink>$$Uhttps://doi.org/10.48550/arXiv.1903.02930$$DView paper in arXiv$$Hfree_for_read</backlink></links><search><creatorcontrib>Anastasopoulos, Antonios</creatorcontrib><creatorcontrib>Kumar, Shankar</creatorcontrib><creatorcontrib>Liao, Hank</creatorcontrib><title>Neural Language Modeling with Visual Features</title><description>Multimodal language models attempt to incorporate non-linguistic features for
the language modeling task. In this work, we extend a standard recurrent neural
network (RNN) language model with features derived from videos. We train our
models on data that is two orders-of-magnitude bigger than datasets used in
prior work. We perform a thorough exploration of model architectures for
combining visual and text features. Our experiments on two corpora (YouCookII
and 20bn-something-something-v2) show that the best performing architecture
consists of middle fusion of visual and text features, yielding over 25%
relative improvement in perplexity. We report analysis that provides insights
into why our multimodal language model improves upon a standard RNN language
model.</description><subject>Computer Science - Computation and Language</subject><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2019</creationdate><recordtype>article</recordtype><sourceid>GOX</sourceid><recordid>eNotzrFuwkAQBNBrKBDkA6jiH7CzvvPZvhIhCJFM0qC01nq965zkADrjJPx9CKGaYkajp9QihSQrrYUnDD_-K0kdmAS0MzBV8SuPAfuowkM3YsfR7thy7w9d9O3PH9G7H8Zru2E8j4GHuZoI9gM_3HOm9pv1frWNq7fnl9WyijEvIC6BxKaGtJQaUJeO8hRdlrNYbOm6cA1RQyJQkGAjwi4nIW6s1hlpa2bq8f_25q1PwX9iuNR_7vrmNr90eT3y</recordid><startdate>20190307</startdate><enddate>20190307</enddate><creator>Anastasopoulos, Antonios</creator><creator>Kumar, Shankar</creator><creator>Liao, Hank</creator><scope>AKY</scope><scope>GOX</scope></search><sort><creationdate>20190307</creationdate><title>Neural Language Modeling with Visual Features</title><author>Anastasopoulos, Antonios ; Kumar, Shankar ; Liao, Hank</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-a670-80cf513c2f820a289c61a946ef5adc6709bccbcff07cfabffe96cfceb5224c253</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2019</creationdate><topic>Computer Science - Computation and Language</topic><toplevel>online_resources</toplevel><creatorcontrib>Anastasopoulos, Antonios</creatorcontrib><creatorcontrib>Kumar, Shankar</creatorcontrib><creatorcontrib>Liao, Hank</creatorcontrib><collection>arXiv Computer Science</collection><collection>arXiv.org</collection></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext_linktorsrc</fulltext></delivery><addata><au>Anastasopoulos, Antonios</au><au>Kumar, Shankar</au><au>Liao, Hank</au><format>journal</format><genre>article</genre><ristype>JOUR</ristype><atitle>Neural Language Modeling with Visual Features</atitle><date>2019-03-07</date><risdate>2019</risdate><abstract>Multimodal language models attempt to incorporate non-linguistic features for
the language modeling task. In this work, we extend a standard recurrent neural
network (RNN) language model with features derived from videos. We train our
models on data that is two orders-of-magnitude bigger than datasets used in
prior work. We perform a thorough exploration of model architectures for
combining visual and text features. Our experiments on two corpora (YouCookII
and 20bn-something-something-v2) show that the best performing architecture
consists of middle fusion of visual and text features, yielding over 25%
relative improvement in perplexity. We report analysis that provides insights
into why our multimodal language model improves upon a standard RNN language
model.</abstract><doi>10.48550/arxiv.1903.02930</doi><oa>free_for_read</oa></addata></record> |
fulltext | fulltext_linktorsrc |
identifier | DOI: 10.48550/arxiv.1903.02930 |
ispartof | |
issn | |
language | eng |
recordid | cdi_arxiv_primary_1903_02930 |
source | arXiv.org |
subjects | Computer Science - Computation and Language |
title | Neural Language Modeling with Visual Features |
url | https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2025-01-08T19%3A48%3A15IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-arxiv_GOX&rft_val_fmt=info:ofi/fmt:kev:mtx:journal&rft.genre=article&rft.atitle=Neural%20Language%20Modeling%20with%20Visual%20Features&rft.au=Anastasopoulos,%20Antonios&rft.date=2019-03-07&rft_id=info:doi/10.48550/arxiv.1903.02930&rft_dat=%3Carxiv_GOX%3E1903_02930%3C/arxiv_GOX%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 |