Human activity recognition (HAR) is one of the important research areas in pervasive computing. Among HAR, sensor-based refers to acquiring a high-level knowledge about human activities from readings many low-level sensor. In recent years, although traditional methods deep learning (DL) have been widely used for HAR with some good performance, they still face such challenges as feature extraction and characterization, continuous action segmentation dealing time series problems. this study,...

Abstract Human Activity Recognition (HAR) is an important research area in human–computer interaction and pervasive computing. In recent years, many deep learning (DL) methods have been widely used for HAR, due to their powerful automatic feature extraction capabilities, they achieve better recognition performance than traditional are applicable more general scenarios. However, the problem that DL increase computational cost of system take up resources while achieving higher accuracy, which...

The global health crisis due to the fast spread of coronavirus disease (Covid-19) has caused great danger all aspects healthcare, economy, and other aspects. highly infectious insidious nature new greatly increases difficulty outbreak prevention control. early rapid detection Covid-19 is an effective way reduce Covid-19. However, detecting accurately quickly in large populations remains be a major challenge worldwide. In this study, A CNN-transformer fusion framework proposed for automatic...

An increasing number of robotic systems contain multiple sensors, and the accurate timestamps sensors are crucial for fusing information in robot applications. However, accuracy stability traditional timestamp synchronization algorithms degraded when synchronizing due to clock inconsistency, asynchronous trigger delay, variable transmission delay. This paper presents a general precise time method with respect local (i.e. typically computer's internal clock). First, reduce software scheduling...

Human activity recognition (HAR) is an important subfield of pervasive computing and pattern recognition. While researchers have achieved remarkable results in feature extraction classification for sensor-based HAR, they encountered performance bottlenecks. Sensor signal denoising has emerged as excellent approach to enhance the HAR architectures. In this study, we propose a novel self-supervised blind method sensor signals, which serves new module task significantly improves overall system...

Since the outbreak of novel coronavirus pneumonia (COVID-19) in 2019, normal learning and living have been severely affected, human life health seriously threatened. Therefore, it is crucial to diagnose rapidly efficiently. In this study, based on classical image classification neural network model, a deep convolutional model attention mechanism proposed named LACNN_CBAM model. The accuracy Acc, precision Pre, recall Rec F-1 scores public dataset collated from published papers are 0.989,...

Even though the RNN, LSTM, and other networks are used to extract dependencies in time series, sensor-based human behavior recognition (HAR) still faces some difficulties, ability of deep learning (DL) features needs be improved. We propose a fusion neural network which an optimized small Convolutional Block Attention Module (MP-CBAM) is suitable for HAR tasks samples. The MP-CBAM added two branches Neural Network (CNN) with different convolution kernel sizes, fused labeled GRU temporal...

Abstract Human Activity Recognition (HAR) is an important research area in human-computer interaction and pervasive computing. In recent years, many deep learning (DL) methods have been widely used for HAR, due to their powerful automatic feature extraction capabilities, they achieve better recognition performance than traditional are applicable more general scenarios. However, the problem that DL increase computational cost of system take up resources while achieving higher accuracy, which...