ML Anthology

Unsupervised Selection of Negative Examples for Grounded Language Learning

Nisha Pillai, Cynthia Matuszek
AAAI 2018 pp. 6517-6523
doi:10.1609/AAAI.V32I1.12108 /aaai/2018/pillai2018aaai-unsupervised/

Abstract

There has been substantial work in recent years on grounded language acquisition, in which language and sensor data are used to create a model relating linguistic constructs to the perceivable world. While powerful, this approach is frequently hindered by ambiguities, redundancies, and omissions found in natural language. We describe an unsupervised system that learns language by training visual classifiers, first selecting important terms from object descriptions, then automatically choosing negative examples from a paired corpus of perceptual and linguistic data. We evaluate the effectiveness of each stage as well as the system's performance on the overall learning task.

PDF AAAI Semantic Scholar

Cite

Text

Pillai and Matuszek. "Unsupervised Selection of Negative Examples for Grounded Language Learning." AAAI Conference on Artificial Intelligence, 2018. doi:10.1609/AAAI.V32I1.12108

Markdown

[Pillai and Matuszek. "Unsupervised Selection of Negative Examples for Grounded Language Learning." AAAI Conference on Artificial Intelligence, 2018.](https://mlanthology.org/aaai/2018/pillai2018aaai-unsupervised/) doi:10.1609/AAAI.V32I1.12108

BibTeX

@inproceedings{pillai2018aaai-unsupervised,
  title     = {{Unsupervised Selection of Negative Examples for Grounded Language Learning}},
  author    = {Pillai, Nisha and Matuszek, Cynthia},
  booktitle = {AAAI Conference on Artificial Intelligence},
  year      = {2018},
  pages     = {6517-6523},
  doi       = {10.1609/AAAI.V32I1.12108},
  url       = {https://mlanthology.org/aaai/2018/pillai2018aaai-unsupervised/}
}