Skip to main content

Research Repository

See what's under the surface

Advanced Search

Zero-shot learning with matching networks for open-ended human activity recognition.

Wijekoon, Anjana; Wiratunga, Nirmalie; Sani, Sadiq

Authors

Anjana Wijekoon

Sadiq Sani



Contributors

Kyle Martin
Editor

Leslie S. Smith
Editor

Abstract

A real-world solution for Human Activity Recognition (HAR) should cover a variety of activities. However training a model to cover each and every possible activity is not practical. Instead we need a solution that can adapt its learning to unseen activities; referred to as open-ended HAR. Recent advancements in deep learning have increasingly begun to focus on the need to learn from few examples, referred to as k-shot learning and to go beyond this to transfer that learning to situations with unseen classes, referred to as zero-shot learning. The latter is particularly relevant to our research in open-ended HAR; and as yet remains unexplored. This paper presents our preliminary work with Zero-shot Learning (ZSL) with a Matching Network to address openended HAR. A Matching Network has the desirable property of learning with few examples and so is well suited to explorations in ZSL. We evaluate Matching Networks for ZSL with a HAR dataset. We propose the use of a variable length support set at test time to overcome the search for the best support set combination that currently plagues the fixed length support set size used by matching nets. Our results show that the variable length approach to be an effective strategy to maintain accuracy whilst avoiding the combinatorial search for the best class combination to form the support set.

Start Date Jun 27, 2018
Publication Date Jul 30, 2018
Print ISSN 1613-0073
Publisher CEUR Workshop Proceedings
Series Title CEUR workshop proceedings
Series Number 2151
Series ISSN 1613-0073
Institution Citation WIJEKOON, A., WIRATUNGA, N. and SANI, S. 2018. Zero-shot learning with matching networks for open-ended human activity recognition. In Martin, K., Wiratunga, N. and Smith, L.S. (eds.) Proceedings of the 2018 Scottish Informatics and Computer Science Alliance (SCISA) workshop on reasoning, learning and explainability (ReaLX 2018), 27 June 2018, Aberdeen, UK. CEUR workshop proceedings, 2151. Aachen: CEUR-WS [online], session 2, paper 4. Available from: http://ceur-ws.org/Vol-2151/Paper_S9.pdf
Keywords Zeroshot learning; Matching networks; Openended HAR
Publisher URL http://ceur-ws.org/Vol-2151/Paper_S9.pdf

Files





You might also like



Downloadable Citations

;