Unsupervised Domain Adaptation for Human Activity Recognition

Paolo Barbosa, Kemilly Dearo Garcia, Joao Mendes-Moreira

Research output: Chapter in Book/Report/Conference proceeding › Chapter › Academic › peer-review

Abstract

Human Activity Recognition has been primarily investigated as a machine learning classification task forcing it to handle with two main limitations. First, it must assume that the testing data has an equal distribution with the training sample. However, the inherent structure of an activity recognition systems is fertile in distribution changes over time, for instance, a specific person can perform physical activities differently from others, and even sensors are prone to misfunction. Secondly, to model the pattern of activities carried out by each user, a significant amount of data is needed. This is impractical especially in the actual era of Big Data with effortless access to public repositories. In order to deal with these problems, this paper investigates the use of Transfer Learning, specifically Unsupervised Domain Adaptation, within human activity recognition systems. The yielded experiment results reveal a useful transfer of knowledge and more importantly the convenience of transfer learning within human activity recognition. Apart from the delineated experiments, our work also contributes to the field of transfer learning in general through an exhaustive survey on transfer learning for human activity recognition based on wearables.
Original languageEnglish
Title of host publicationIntelligent Data Engineering and Automated Learning
Subtitle of host publicationIDEAL 2018
PublisherSpringer
Pages623-630
ISBN (Electronic)978-3-030-03493-1
ISBN (Print)978-3-030-03492-4
Publication statusPublished - 21 Nov 2018

Publication series

NameLecture Notes in Computer Science
PublisherSpringer, Cham
Volume11314
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Fingerprint

Unsupervised learning
Learning systems
Experiments
Sensors
Testing
Big data

Cite this

Barbosa, P., Dearo Garcia, K., & Mendes-Moreira, J. (2018). Unsupervised Domain Adaptation for Human Activity Recognition. In Intelligent Data Engineering and Automated Learning: IDEAL 2018 (pp. 623-630). (Lecture Notes in Computer Science; Vol. 11314). Springer.
Barbosa, Paolo ; Dearo Garcia, Kemilly ; Mendes-Moreira, Joao. / Unsupervised Domain Adaptation for Human Activity Recognition. Intelligent Data Engineering and Automated Learning: IDEAL 2018. Springer, 2018. pp. 623-630 (Lecture Notes in Computer Science).
@inbook{8024fa4c5e034c5f93436c3efece6e01,
title = "Unsupervised Domain Adaptation for Human Activity Recognition",
abstract = "Human Activity Recognition has been primarily investigated as a machine learning classification task forcing it to handle with two main limitations. First, it must assume that the testing data has an equal distribution with the training sample. However, the inherent structure of an activity recognition systems is fertile in distribution changes over time, for instance, a specific person can perform physical activities differently from others, and even sensors are prone to misfunction. Secondly, to model the pattern of activities carried out by each user, a significant amount of data is needed. This is impractical especially in the actual era of Big Data with effortless access to public repositories. In order to deal with these problems, this paper investigates the use of Transfer Learning, specifically Unsupervised Domain Adaptation, within human activity recognition systems. The yielded experiment results reveal a useful transfer of knowledge and more importantly the convenience of transfer learning within human activity recognition. Apart from the delineated experiments, our work also contributes to the field of transfer learning in general through an exhaustive survey on transfer learning for human activity recognition based on wearables.",
author = "Paolo Barbosa and {Dearo Garcia}, Kemilly and Joao Mendes-Moreira",
year = "2018",
month = "11",
day = "21",
language = "English",
isbn = "978-3-030-03492-4",
series = "Lecture Notes in Computer Science",
publisher = "Springer",
pages = "623--630",
booktitle = "Intelligent Data Engineering and Automated Learning",

}

Barbosa, P, Dearo Garcia, K & Mendes-Moreira, J 2018, Unsupervised Domain Adaptation for Human Activity Recognition. in Intelligent Data Engineering and Automated Learning: IDEAL 2018. Lecture Notes in Computer Science, vol. 11314, Springer, pp. 623-630.

Unsupervised Domain Adaptation for Human Activity Recognition. / Barbosa, Paolo; Dearo Garcia, Kemilly ; Mendes-Moreira, Joao.

Intelligent Data Engineering and Automated Learning: IDEAL 2018. Springer, 2018. p. 623-630 (Lecture Notes in Computer Science; Vol. 11314).

Research output: Chapter in Book/Report/Conference proceeding › Chapter › Academic › peer-review

TY - CHAP

T1 - Unsupervised Domain Adaptation for Human Activity Recognition

AU - Barbosa, Paolo

AU - Dearo Garcia, Kemilly

AU - Mendes-Moreira, Joao

PY - 2018/11/21

Y1 - 2018/11/21

N2 - Human Activity Recognition has been primarily investigated as a machine learning classification task forcing it to handle with two main limitations. First, it must assume that the testing data has an equal distribution with the training sample. However, the inherent structure of an activity recognition systems is fertile in distribution changes over time, for instance, a specific person can perform physical activities differently from others, and even sensors are prone to misfunction. Secondly, to model the pattern of activities carried out by each user, a significant amount of data is needed. This is impractical especially in the actual era of Big Data with effortless access to public repositories. In order to deal with these problems, this paper investigates the use of Transfer Learning, specifically Unsupervised Domain Adaptation, within human activity recognition systems. The yielded experiment results reveal a useful transfer of knowledge and more importantly the convenience of transfer learning within human activity recognition. Apart from the delineated experiments, our work also contributes to the field of transfer learning in general through an exhaustive survey on transfer learning for human activity recognition based on wearables.

AB - Human Activity Recognition has been primarily investigated as a machine learning classification task forcing it to handle with two main limitations. First, it must assume that the testing data has an equal distribution with the training sample. However, the inherent structure of an activity recognition systems is fertile in distribution changes over time, for instance, a specific person can perform physical activities differently from others, and even sensors are prone to misfunction. Secondly, to model the pattern of activities carried out by each user, a significant amount of data is needed. This is impractical especially in the actual era of Big Data with effortless access to public repositories. In order to deal with these problems, this paper investigates the use of Transfer Learning, specifically Unsupervised Domain Adaptation, within human activity recognition systems. The yielded experiment results reveal a useful transfer of knowledge and more importantly the convenience of transfer learning within human activity recognition. Apart from the delineated experiments, our work also contributes to the field of transfer learning in general through an exhaustive survey on transfer learning for human activity recognition based on wearables.

M3 - Chapter

SN - 978-3-030-03492-4

T3 - Lecture Notes in Computer Science

SP - 623

EP - 630

BT - Intelligent Data Engineering and Automated Learning

PB - Springer

ER -

Barbosa P, Dearo Garcia K, Mendes-Moreira J. Unsupervised Domain Adaptation for Human Activity Recognition. In Intelligent Data Engineering and Automated Learning: IDEAL 2018. Springer. 2018. p. 623-630. (Lecture Notes in Computer Science).