Articles and book chapters Articles and book chapters

Return to Full Page

Humans meet models on object naming: a new dataset and analysis

  • Authors
  • Silberer C, Zarrieß S, Westera M, Boleda G
  • UPF authors
  • BOLEDA TORRENT, GEMMA; WESTERA ., MATTHIJS; SILBERER ., CARINA HELGA;
  • Authors of the book
  • Scott D, Bel N, Zong C
  • Book title
  • apos;2020)
  • Publisher
  • International Committee on Computational Linguistics
  • Publication year
  • 2020
  • Pages
  • 1893-1905
  • ISBN
  • 978-1-952148-27-9
  • Abstract
  • amp;}V model and demonstrate its effectiveness on the task of object naming. However, our fine-grained analysis reveals that what appears to be human-like model behavior is not stable across domains, e.g., the model confuses people and clothing objects much more frequently than humans do. We also find that standard evaluations underestimate the actual effectiveness of the naming model: on the single-label names of the original dataset (Visual Genome), it obtains â¿¿27{\%} accuracy points than on MN v2, that includes all valid object names.
  • Complete citation
  • Silberer C, Zarrieß S, Westera M, Boleda G. Humans meet models on object naming: a new dataset and analysis. In: Scott D, Bel N, Zong C. Proceedings of the 28th International Conference on Computational Linguistics (COLING'2020). 1 ed. International Committee on Computational Linguistics; 2020. p. 1893-1905.