Comparatives, quantifiers, proportions: a multi-task model for the learning of quantities from vision

Sorodoc, Ionut-Teodor; Pezzelle, Sandro; Bernardi, Raffaella

Comparatives, quantifiers, proportions: a multi-task model for the learning of quantities from vision

License

Citation

Pezzelle S, Sorodoc IT, Bernardi R. Comparatives, quantifiers, proportions: a multi-task model for the learning of quantities from vision. In: Proceedings of the 2018 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies; 2018 Jun 1-6; New Orleans, United States of America. Stroudsburg (PA): ACL; 2018. p. 419-30.

Abstract

The present work investigates whether different quantification mechanisms (set comparison, vague quantification, and proportional estimation) can be jointly learned from visual scenes by a multi-task computational model. The motivation is that, in humans, these processes underlie the same cognitive, nonsymbolic ability, which allows an automatic estimation and comparison of set magnitudes. We show that when information about lowercomplexity tasks is available, the higher-level proportional task becomes more accurate than when performed in isolation. Moreover, the multi-task model is able to generalize to unseen combinations of target/non-target objects. Consistently with behavioral evidence showing the interference of absolute number in the proportional task, the multi-task model no longer works when asked to provide the number of target objects in the scene.

Description

Comunicació presentada a la Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (NAACL-HLT 2018), celebrada els dies 1 a 6 de juny de 2018 a Nova Orleans, Estats Units d'Amèrica.