VisualSem: A High-quality Knowledge Graph for Vision & Language

Houda Alberts, Ningyuan Huang, Yash R. Deshpande, Yibo Liu, Kyunghyun Cho, Clara Vania, Iacer Calixto

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

An exciting frontier in natural language understanding (NLU) and generation (NLG) calls for (vision-and-) language models that can efficiently access external structured knowledge repositories. However, many existing knowledge bases only cover limited domains, or suffer from noisy data, and most of all are typically hard to integrate into neural language pipelines. To fill this gap, we release VisualSem: a high-quality knowledge graph (KG) which includes nodes with multilingual glosses, multiple illustrative images, and visually relevant relations. We also release a neural multi-modal retrieval model that can use images or sentences as inputs and retrieves entities in the KG. This multi-modal retrieval model can be integrated into any (neural network) model pipeline. We encourage the research community to use VisualSem for data augmentation and/or as a source of grounding, among other possible uses. VisualSem as well as the multi-modal retrieval models are publicly available and can be downloaded in this URL: https://github.com/iacercalixto/visualsem.

Original languageEnglish (US)
Title of host publicationMRL 2021 - 1st Workshop on Multilingual Representation Learning, Proceedings of the Conference
EditorsDuygu Ataman, Alexandra Birch, Alexis Conneau, Orhan Firat, Sebastian Ruder, Gozde Gul Sahin
PublisherAssociation for Computational Linguistics (ACL)
Pages138-152
Number of pages15
ISBN (Electronic)9781954085961
StatePublished - 2021
Event1st Workshop on Multilingual Representation Learning, MRL 2021 - Punta Cana, Dominican Republic
Duration: Nov 11 2021 → …

Publication series

NameMRL 2021 - 1st Workshop on Multilingual Representation Learning, Proceedings of the Conference

Conference

Conference1st Workshop on Multilingual Representation Learning, MRL 2021
Country/TerritoryDominican Republic
CityPunta Cana
Period11/11/21 → …

ASJC Scopus subject areas

  • Linguistics and Language
  • Language and Linguistics

Fingerprint

Dive into the research topics of 'VisualSem: A High-quality Knowledge Graph for Vision & Language'. Together they form a unique fingerprint.

Cite this