@inproceedings{jolly-kapoor-2020-pre,
title = "Can Pre-training help {VQA} with Lexical Variations?",
author = "Jolly, Shailza and
Kapoor, Shubham",
editor = "Cohn, Trevor and
He, Yulan and
Liu, Yang",
booktitle = "Findings of the Association for Computational Linguistics: EMNLP 2020",
month = nov,
year = "2020",
address = "Online",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2020.findings-emnlp.257",
doi = "10.18653/v1/2020.findings-emnlp.257",
pages = "2863--2868",
abstract = "Rephrasings or paraphrases are sentences with similar meanings expressed in different ways. Visual Question Answering (VQA) models are closing the gap with the oracle performance for datasets like VQA2.0. However, these models fail to perform well on rephrasings of a question, which raises some important questions like Are these models robust towards linguistic variations? Is it the architecture or the dataset that we need to optimize? In this paper, we analyzed VQA models in the space of paraphrasing. We explored the role of language {\&} cross-modal pre-training to investigate the robustness of VQA models towards lexical variations. Our experiments find that pre-trained language encoders generate efficient representations of question rephrasings, which help VQA models correctly infer these samples. We empirically determine why pre-training language encoders improve lexical robustness. Finally, we observe that although pre-training all VQA components obtain state-of-the-art results on the VQA-Rephrasings dataset, it still fails to completely close the performance gap between original and rephrasing validation splits.",
}
<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="jolly-kapoor-2020-pre">
<titleInfo>
<title>Can Pre-training help VQA with Lexical Variations?</title>
</titleInfo>
<name type="personal">
<namePart type="given">Shailza</namePart>
<namePart type="family">Jolly</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Shubham</namePart>
<namePart type="family">Kapoor</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2020-11</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Findings of the Association for Computational Linguistics: EMNLP 2020</title>
</titleInfo>
<name type="personal">
<namePart type="given">Trevor</namePart>
<namePart type="family">Cohn</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yulan</namePart>
<namePart type="family">He</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yang</namePart>
<namePart type="family">Liu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Online</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
</relatedItem>
<abstract>Rephrasings or paraphrases are sentences with similar meanings expressed in different ways. Visual Question Answering (VQA) models are closing the gap with the oracle performance for datasets like VQA2.0. However, these models fail to perform well on rephrasings of a question, which raises some important questions like Are these models robust towards linguistic variations? Is it the architecture or the dataset that we need to optimize? In this paper, we analyzed VQA models in the space of paraphrasing. We explored the role of language & cross-modal pre-training to investigate the robustness of VQA models towards lexical variations. Our experiments find that pre-trained language encoders generate efficient representations of question rephrasings, which help VQA models correctly infer these samples. We empirically determine why pre-training language encoders improve lexical robustness. Finally, we observe that although pre-training all VQA components obtain state-of-the-art results on the VQA-Rephrasings dataset, it still fails to completely close the performance gap between original and rephrasing validation splits.</abstract>
<identifier type="citekey">jolly-kapoor-2020-pre</identifier>
<identifier type="doi">10.18653/v1/2020.findings-emnlp.257</identifier>
<location>
<url>https://aclanthology.org/2020.findings-emnlp.257</url>
</location>
<part>
<date>2020-11</date>
<extent unit="page">
<start>2863</start>
<end>2868</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Can Pre-training help VQA with Lexical Variations?
%A Jolly, Shailza
%A Kapoor, Shubham
%Y Cohn, Trevor
%Y He, Yulan
%Y Liu, Yang
%S Findings of the Association for Computational Linguistics: EMNLP 2020
%D 2020
%8 November
%I Association for Computational Linguistics
%C Online
%F jolly-kapoor-2020-pre
%X Rephrasings or paraphrases are sentences with similar meanings expressed in different ways. Visual Question Answering (VQA) models are closing the gap with the oracle performance for datasets like VQA2.0. However, these models fail to perform well on rephrasings of a question, which raises some important questions like Are these models robust towards linguistic variations? Is it the architecture or the dataset that we need to optimize? In this paper, we analyzed VQA models in the space of paraphrasing. We explored the role of language & cross-modal pre-training to investigate the robustness of VQA models towards lexical variations. Our experiments find that pre-trained language encoders generate efficient representations of question rephrasings, which help VQA models correctly infer these samples. We empirically determine why pre-training language encoders improve lexical robustness. Finally, we observe that although pre-training all VQA components obtain state-of-the-art results on the VQA-Rephrasings dataset, it still fails to completely close the performance gap between original and rephrasing validation splits.
%R 10.18653/v1/2020.findings-emnlp.257
%U https://aclanthology.org/2020.findings-emnlp.257
%U https://doi.org/10.18653/v1/2020.findings-emnlp.257
%P 2863-2868
Markdown (Informal)
[Can Pre-training help VQA with Lexical Variations?](https://aclanthology.org/2020.findings-emnlp.257) (Jolly & Kapoor, Findings 2020)
ACL