Abstract
With the development of the Internet of things (IoT) and wearable devices, the sensor-based human activity recognition (HAR) has attracted more and more attentions from researchers due to its outstanding characteristics of convenience and privacy. Meanwhile, deep learning algorithms can extract high-dimensional features automatically, which makes it possible to achieve the end-to-end learning. Especially the convolutional neural network (CNN) has been widely used in the field of computer vision, while the influence of environmental background, camera shielding, and other factors are the biggest challenges to it. However, the sensor-based HAR can circumvent these problems well. Two improved HAR methods based on Gramian angular field (GAF) and deep CNN are proposed in this paper. Firstly, the GAF algorithm is used to transform the one-dimensional sensor data into the two-dimensional images. Then, through the multi-dilated kernel residual (Mdk-Res) module, a new improved deep CNN network Mdk-ResNet is proposed, which extracts the features among sampling points with different intervals. Furthermore, the Fusion-Mdk-ResNet is adopted to process and fuse data collected by different sensors automatically. The comparative experiments are conducted on three public activity datasets, which are WISDM, UCI HAR and OPPORTUNITY. The optimal results are obtained by using the indexes such as accuracy, precision, recall and F-measure, which verifies the effectiveness of the proposed methods.
Highlights
With the rapid development of the 5th generation (5G) mobile networks, Internet of things (IoT) and artificial intelligence (AI), the technology of human activity recognition (HAR) is becoming more and more important in people's daily lives because of its ability to analyze and recognize human activities by the raw sensor data
The results showed that the deep convolutional neural network (CNN) using dropout got better recognition performance, and the training time of deep CNN was much less than that of long short-term memory (LSTM)
The OPPORTUNITY dataset contains data collected by multiple sensors, which are classified by using the FusionMdk-ResNet
Summary
With the rapid development of the 5th generation (5G) mobile networks, Internet of things (IoT) and artificial intelligence (AI), the technology of human activity recognition (HAR) is becoming more and more important in people's daily lives because of its ability to analyze and recognize human activities by the raw sensor data. The experimental results show that the new HAR method proposed in this paper can effectively improve the multi-scale feature extraction capability and the accuracy of activity recognition by combining the characteristics of GAF algorithm, the structure and advantages of CNN, residual learning and dilated convolution. (2) A new improved deep CNN network Mdk-ResNet based on the multi-dilated kernel residual (Mdk-Res) modules is proposed in this paper, which makes it possible to extract rich features among the sampling points with different time intervals, thereby improving the recognition accuracy. The Mdk-Res module uses multiple normal convolution kernels and dilated convolution kernels at the same time, which improves the ability of the network to extract features of different scales. ResNet, firstly, the one-dimensional time series collected by multiple sensors is transformed into the two-dimensional images through the GAF algorithm, and features are extracted by multiple Mdk-Res modules, respectively.
Sign-in/Register to view highlights & summary 
Published Version (
Free)
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call