Leveraging Past References for Robust Language Grounding
Name
K19-1040.pdf
Description
Published version
Size
4.45 MB
Format
Adobe PDF
Checksum (MD5)
6e6da6df27c515369c5c5c1e66f7ac05
Author(s)
Roy, Subhro
Noseworthy, Michael
Paul, Rohan
Park, Daehyung
Roy, Nicholas
Date Issued
November 2019
Journal
CoNLL 2019 - 23rd Conference on Computational Natural Language Learning, Proceedings of the Conference
Publisher
Association for Computational Linguistics (ACL)
Citation
Roy, Subhro, Noseworthy, Michael, Paul, Rohan, Park, Daehyung and Roy, Nicholas. 2019. "Leveraging Past References for Robust Language Grounding." CoNLL 2019 - 23rd Conference on Computational Natural Language Learning, Proceedings of the Conference.
Version
Final published version
Abstract
© 2019 Association for Computational Linguistics. Grounding referring expressions to objects in an environment has traditionally been considered a one-off, ahistorical task. However, in realistic applications of grounding, multiple users will repeatedly refer to the same set of objects. As a result, past referring expressions for objects can provide strong signals for grounding subsequent referring expressions. We therefore reframe the grounding problem from the perspective of coreference detection and propose a neural network that detects when two expressions are referring to the same object. The network combines information from vision and past referring expressions to resolve which object is being referred to. Our experiments show that detecting referring expression coreference is an effective way to ground objects described by subtle visual properties, which standard visual grounding models have difficulty capturing. We also show the ability to detect object coreference allows the grounding model to perform well even when it encounters object categories not seen in the training data.
MIT Department
Massachusetts Institute of Technology. Computer Science and Artificial Intelligence Laboratory
Terms of Use
Article is made available in accordance with the publisher's policy and may be subject to US copyright law. Please refer to the publisher's site for terms of use.
Persistent DSpace Link
DOI of Published Version
https://doi.org/10.18653/V1/K19-1040
