Attentive recurrent comparators

Shyam, P and Gupta, S and Dukkipati, A (2017) Attentive recurrent comparators. In: 34th International Conference on Machine Learning, ICML 2017, 6 - 11 August 2017, Sydney, pp. 4890-4898.

PDF ICML_7-4890 - 4898_2017.pdf - Published Version Restricted to Registered users only Download (556kB) | Request a copy

Abstract

Rapid learning requires flexible representations to quickly adopt to new evidence. Wc develop a novel class of models called Attentive Recurrent Comparators (ARCs) that form representations of objects by cycling through them and making observations. Using the representations extracted by ARCs, we develop a way of approximating a dynamic representation space and use it for one-shot learning. In the task of one-shot classification on the Omniglot dataset, we achieve the state of the art performance with an error rate of 1.5. This represents the first super-human result achieved for this task with a generic model that uses only pixel information.

Item Type:	Conference Paper
Publication:	34th International Conference on Machine Learning, ICML 2017
Publisher:	International Machine Learning Society (IMLS)
Additional Information:	The copyright for this article belongs to International Machine Learning Society (IMLS)
Keywords:	Artificial intelligence; Classification (of information); Comparators (optical); Learning systems, Dynamic representation; Error rate; Generic modeling; One-shot learning; Pixel information; Rapid learning; Shot classification; State-of-the-art performance, Comparator circuits
Department/Centre:	Division of Electrical Sciences > Computer Science & Automation
Date Deposited:	27 Jul 2022 05:25
Last Modified:	27 Jul 2022 05:25
URI:	https://eprints.iisc.ac.in/id/eprint/74676

Actions (login required)

View Item