Iss2Image: A Novel Signal-Encoding Technique for CNN-Based Human Activity Recognition
| Autor: | Thien Huynh-The, Jongwon Lee, Jee-In Kim, Sungyoung Lee, Tae Ho Hur, Jae Hun Bang |
|---|---|
| Jazyk: | angličtina |
| Rok vydání: | 2018 |
| Předmět: |
Adult
Male human activity recognition Computer science convolutional neural network Biosensing Techniques 02 engineering and technology Accelerometer lcsh:Chemical technology smartphone Biochemistry Convolutional neural network Signal Article smartwatch Pattern Recognition Automated Analytical Chemistry Activity recognition Distortion Accelerometry 0202 electrical engineering electronic engineering information engineering Humans lcsh:TP1-1185 Electrical and Electronic Engineering Exercise Instrumentation business.industry Deep learning 020206 networking & telecommunications Pattern recognition signal transformation Middle Aged Scale invariance Atomic and Molecular Physics and Optics encoder accelerometer Female 020201 artificial intelligence & image processing Neural Networks Computer Artificial intelligence business Encoder |
| Zdroj: | Sensors Volume 18 Issue 11 Sensors, Vol 18, Iss 11, p 3910 (2018) Sensors (Basel, Switzerland) |
| ISSN: | 1424-8220 |
| DOI: | 10.3390/s18113910 |
| Popis: | The most significant barrier to success in human activity recognition is extracting and selecting the right features. In traditional methods, the features are chosen by humans, which requires the user to have expert knowledge or to do a large amount of empirical study. Newly developed deep learning technology can automatically extract and select features. Among the various deep learning methods, convolutional neural networks (CNNs) have the advantages of local dependency and scale invariance and are suitable for temporal data such as accelerometer (ACC) signals. In this paper, we propose an efficient human activity recognition method, namely Iss2Image (Inertial sensor signal to Image), a novel encoding technique for transforming an inertial sensor signal into an image with minimum distortion and a CNN model for image-based activity classification. Iss2Image converts real number values from the X, Y, and Z axes into three color channels to precisely infer correlations among successive sensor signal values in three different dimensions. We experimentally evaluated our method using several well-known datasets and our own dataset collected from a smartphone and smartwatch. The proposed method shows higher accuracy than other state-of-the-art approaches on the tested datasets. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|
| Nepřihlášeným uživatelům se plný text nezobrazuje | K zobrazení výsledku je třeba se přihlásit. |