Intelligent system for human activity recognition in IoT environment

Abstract

In recent years, the adoption of machine learning has grown steadily in different fields affecting the day-to-day decisions of individuals. This paper presents an intelligent system for recognizing human’s daily activities in a complex IoT environment. An enhanced model of capsule neural network called 1D-HARCapsNe is proposed. This proposed model consists of convolution layer, primary capsule layer, activity capsules flat layer and output layer. It is validated using WISDM dataset collected via smart devices and normalized using the random-SMOTE algorithm to handle the imbalanced behavior of the dataset. The experimental results indicate the potential and strengths of the proposed 1D-HARCapsNet that achieved enhanced performance with an accuracy of 98.67%, precision of 98.66%, recall of 98.67%, and F1-measure of 0.987 which shows major performance enhancement compared to the Conventional CapsNet (accuracy 90.11%, precision 91.88%, recall 89.94%, and F1-measure 0.93).

Human Activity Recognition Based on a Modified Capsule Network

“…Table 8 lists the methods and results of some other studies on WISDM. Among them, CapsNet is also applied for HAR in references [29][30][31]. Te accuracy is higher than references [29,31] but slightly lower than reference [30].…”

“…Among them, CapsNet is also applied for HAR in references [29][30][31]. Te accuracy is higher than references [29,31] but slightly lower than reference [30]. Because the WISDM dataset is a very imbalanced dataset, the random SMOTE algorithm was employed in reference [30] to handle the imbalanced issue of the dataset, which is more conducive to training the network model.…”

“…Te accuracy is higher than references [29,31] but slightly lower than reference [30]. Because the WISDM dataset is a very imbalanced dataset, the random SMOTE algorithm was employed in reference [30] to handle the imbalanced issue of the dataset, which is more conducive to training the network model. Table 8 also shows that the MCN model proposed in this study outperforms CNN, LSTM, and their combination on WISDM.…”

“…Te results demonstrated that the method achieved a higher accuracy than CNN and recurrent neural network (RNN). Khaled et al [30] proposed an enhanced model of CapsNet named 1D-HARCapsNe. Tis provided an efcient intelligent decision-support approach for HAR.…”

A privacy-preserving student status monitoring system

“…Increasingly, computers are auxiliary system for many scenes; part of these systems require the participation of record equipment such as cameras [1][2][3]. Ubiquitous camera means that they have access or record to amount of data; much of them are private in nature.…”