University of Sussex
Browse
2016 - J - Ordonez - Deep Convolutional and LSTM Recurrent Neural Networks for Multimodal Wearable Activity Recognition (Sensors, 2016).pdf (2.11 MB)

Deep convolutional and LSTM recurrent neural networks for multimodal wearable activity recognition

Download (2.11 MB)
journal contribution
posted on 2023-06-09, 00:03 authored by Francisco Javier Ordonez Morales, Daniel RoggenDaniel Roggen
Human activity recognition (HAR) tasks have traditionally been solved using engineered features obtained by heuristic processes. Current research suggests that deep convolutional neural networks are suited to automate feature extraction from raw sensor inputs. However, human activities are made of complex sequences of motor movements, and capturing this temporal dynamics is fundamental for successful HAR. Based on the recent success of recurrent neural networks for time series domains, we propose a generic deep framework for activity recognition based on convolutional and LSTM recurrent units, which: (i) is suitable for multimodal wearable sensors; (ii) can perform sensor fusion naturally; (iii) does not require expert knowledge in designing features; and (iv) explicitly models the temporal dynamics of feature activations. We evaluate our framework on two datasets, one of which has been used in a public activity recognition challenge. Our results show that our framework outperforms competing deep non-recurrent networks on the challenge dataset by 4% on average; outperforming some of the previous reported results by up to 9%. Our results show that the framework can be applied to homogeneous sensor modalities, but can also fuse multimodal sensors to improve performance. We characterise key architectural hyperparameters’ influence on performance to provide insights about their optimisation.

Funding

Lifelearn: Unbounded activity and context awareness; G1786; EPSRC-ENGINEERING & PHYSICAL SCIENCES RESEARCH COUNCIL; EP/N007816/1

Is deep learning useful for wearable activity recognition?; G1460; GOOGLE

History

Publication status

  • Published

File Version

  • Published version

Journal

Sensors

ISSN

1424-8220

Publisher

MDPI

Issue

1

Volume

16

Page range

1-25

Department affiliated with

  • Engineering and Design Publications

Full text available

  • Yes

Peer reviewed?

  • Yes

Legacy Posted Date

2016-01-18

First Open Access (FOA) Date

2016-01-18

First Compliant Deposit (FCD) Date

2016-01-18

Usage metrics

    University of Sussex (Publications)

    Categories

    No categories selected

    Licence

    Exports

    RefWorks
    BibTeX
    Ref. manager
    Endnote
    DataCite
    NLM
    DC