Grounding language acquisition by training semantic parsersusing captioned videos

TitleGrounding language acquisition by training semantic parsersusing captioned videos
Publication TypeConference Paper
Year of Publication2018
AuthorsRoss, C, Barbu, A, Berzak, Y, Myanganbayar, B, Katz, B
Conference NameProceedings of the 2018 Conference on Empirical Methods in Natural Language Processing (EMNLP 2018),
Date Published10/2018
Conference LocationBrussels, Belgium
ISBN Number978-1-948087-84-1

We develop a semantic parser that is trained ina grounded setting using pairs of videos cap-tioned with sentences. This setting is bothdata-efficient, requiring little annotation, andsimilar to the experience of children wherethey observe their environment and listen tospeakers. The semantic parser recovers themeaning of English sentences despite not hav-ing access to any annotated sentences. It doesso despite the ambiguity inherent in visionwhere a sentence may refer to any combina-tion of objects, object properties, relations oractions taken by any agent in a video. For thistask, we collected a new dataset for groundedlanguage acquisition. Learning a grounded se-mantic parser — turning sentences into logi-cal forms using captioned videos — can sig-nificantly expand the range of data that parserscan be trained on, lower the effort of training asemantic parser, and ultimately lead to a betterunderstanding of child language acquisition.

Download:  PDF icon Ross-et-al_ACL2018_Grounding language acquisition by training semantic parsing using caption videos.pdf

Associated Module: 

Research Area: 

CBMM Relationship: 

  • CBMM Funded