A Human Activity Recognition Method Based on Lightweight Feature Extraction Combined with Pruned and Quantized CNN for Wearable Device
- Authors
- Yi, Myung-Kyu; Lee, Wai-Kong; Hwang, Seong Oun
- Issue Date
- Aug-2023
- Publisher
- Institute of Electrical and Electronics Engineers Inc.
- Keywords
- Accelerometers; Biomedical monitoring; convolutional neural network; Convolutional neural networks; deep learning; Feature extraction; human activity recognition; Support vector machines; Wearable computers; wearable sensors; Wearable sensors
- Citation
- IEEE Transactions on Consumer Electronics, v.69, no.3, pp.657 - 670
- Journal Title
- IEEE Transactions on Consumer Electronics
- Volume
- 69
- Number
- 3
- Start Page
- 657
- End Page
- 670
- URI
- https://scholarworks.bwise.kr/gachon/handle/2020.sw.gachon/89021
- DOI
- 10.1109/TCE.2023.3266506
- ISSN
- 0098-3063
- Abstract
- Human Activity Recognition (HAR) is becoming an essential part of human life care. Existing HAR methods are usually developed using a two-level approach, wherein a first-level Machine Learning (ML) classifier is employed to distinguish the static and dynamic activities, followed by a second-level classifier to identify the specific activity. These approaches are not suitable for wearable devices, due to the high computational and memory consumption. Our rigorous analysis of various HAR datasets opens up a new possibility that static or dynamic activities can be discriminated against through a simple statistical technique. Therefore, we propose to utilize a statistical feature extraction technique to replace the first-level ML classifier, thus achieving more lightweight computation. Next, we employ Random Forest (RF) and Convolutional Neural Networks (CNN) to classify the specific activities, achieving higher accuracy compared to the state-of-the-art results. We further reduce the computation and memory consumption of the above combined approach by applying pruning and quantizing techniques to CNN (PQ-CNN). Experimental results show the proposed lightweight HAR method achieved an F1 score of 0.9417 and 0.9438 for unbalanced and balanced datasets, respectively. On top of lightweight and accuracy, the proposed HAR method is practical for wearable devices by using a single accelerometer. IEEE
- Files in This Item
- There are no files associated with this item.
- Appears in
Collections - IT융합대학 > 컴퓨터공학과 > 1. Journal Articles
![qrcode](https://api.qrserver.com/v1/create-qr-code/?size=55x55&data=https://scholarworks.bwise.kr/gachon/handle/2020.sw.gachon/89021)
Items in ScholarWorks are protected by copyright, with all rights reserved, unless otherwise indicated.