Elderly population (over the age of 60) is predicted to be 1.2 billion by 2025. Most of the elderly people would like to stay alone in their own house due to the high eldercare cost and privacy invasion. Unobtrusive activity recognition is the most preferred solution for monitoring daily activities of the elderly people living alone rather than the camera and wearable devices based systems. Thus, we propose an unobtrusive activity recognition classifier using Deep Convolutional Neural Network (DCNN) and anonymous binary sensors (PIR sensors and door sensors). The real world long-term fully annotated Aruba open dataset collected by binary sensors has been employed for the training and evaluation of the classifier. First, ten basic daily activities, namely: Bed_to_Toilet, Eating, Meal_Preparation, Relax, Sleeping, Work, Housekeeping, Wash_Dishes, Enter_Home, and Leave_Home are segmented with different sliding window sizes, and then converted into binary activity images. Next, the activity images are employed for training the DCNN classifiers with different parameters. Finally, the trained classifiers are evaluated with the 10-fold cross validation method, and the results show that the best DCNN classifier gives 0.79 and 0.951 of F1-score for all 10 activities and eight activities (excluding Leave_Home and Wash_Dishes), respectively.

A reported 30% of people worldwide have abnormal lung sounds, including crackles, rhonchi, and wheezes. To date, the traditional stethoscope remains the most popular tool used by physicians to diagnose such abnormal lung sounds, however, many problems arise with the use of a stethoscope, including the effects of environmental noise, the inability to record and store lung sounds for follow-up or tracking, and the physician’s subjective diagnostic experience. This study has developed a digital stethoscope to help physicians overcome these problems when diagnosing abnormal lung sounds. In this digital system, mel-frequency cepstral coefficients (MFCCs) were used to extract the features of lung sounds, and then the K-means algorithm was used for feature clustering, to reduce the amount of data for computation. Finally, the K-nearest neighbor method was used to classify the lung sounds. The proposed system can also be used for home care: if the percentage of abnormal lung sound frames is > 30% of the whole test signal, the system can automatically warn the user to visit a physician for diagnosis. We also used bend sensors together with an amplification circuit, Bluetooth, and a microcontroller to implement a respiration detector. The respiratory signal extracted by the bend sensors can be transmitted to the computer via Bluetooth to calculate the respiratory cycle, for real-time assessment. If an abnormal status is detected, the device will warn the user automatically. Experimental results indicated that the error in respiratory cycles between measured and actual values was only 6.8%, illustrating the potential of our detector for home care applications.

Human activity recognition (HAR) can monitor persons at risk of COVID-19 virus infection to manage their activity status. Currently, many people are isolated at home or quarantined in some specified places due to the spread of COVID-19 virus all over the world. This situation raises the requirement of using the HAR to observe physical activity levels to assess physical and mental health. This study proposes an ensemble learning algorithm (ELA) to perform activity recognition using the signals recorded by smartphone sensors. The proposed ELA combines a gated recurrent unit (GRU), a convolutional neural network (CNN) stacked on the GRU and a deep neural network (DNN). The input samples of DNN were an extra feature vector consisting of 561 time-domain and frequency-domain parameters. The full connected DNN was used to fuse three models for the activity classification. The experimental results show that the precision, recall, F1-score and accuracy achieved by the ELA are 96.8%, 96.8%, 96.8%, and 96.7%, respectively, which are superior to the existing schemes.

The performance of hyperspectral image (HSI) classification is highly dependent on spatial and spectral information, and is heavily affected by factors such as data redundancy and insufficient spatial resolution. To overcome these challenges, many convolutional neural networks (CNN) especially 2D-CNN-based methods have been proposed for HSI classification. However, these methods produced insufficient results compared to 3D-CNN-based methods. On the other hand, the high computational complexity of the 3D-CNN-based methods is still a major concern that needs to be addressed. Therefore, this study introduces a consolidated convolutional neural network (C-CNN) to overcome the aforementioned issues. The proposed C-CNN is comprised of a three-dimension CNN (3D-CNN) joined with a two-dimension CNN (2D-CNN). The 3D-CNN is used to represent spatial–spectral features from the spectral bands, and the 2D-CNN is used to learn abstract spatial features. Principal component analysis (PCA) was firstly applied to the original HSIs before they are fed to the network to reduce the spectral bands redundancy. Moreover, image augmentation techniques including rotation and flipping have been used to increase the number of training samples and reduce the impact of overfitting. The proposed C-CNN that was trained using the augmented images is named C-CNN-Aug. Additionally, both Dropout and L2 regularization techniques have been used to further reduce the model complexity and prevent overfitting. The experimental results proved that the proposed model can provide the optimal trade-off between accuracy and computational time compared to other related methods using the Indian Pines, Pavia University, and Salinas Scene hyperspectral benchmark datasets.

Many issues have been identified in outdoor teaching, especially in places that lack the capacity to effectively present information about such subjects as historical relics, rare animals, and geological landscapes. This study proposes an Environment of Ubiquitous Learning with Educational Resources (EULER) based on radio frequency identification (RFID), the Internet, ubiquitous computing, embedded systems and database technologies to resolve such issues. A case study of natural science learning was conducted in classrooms and at the Guandu Nature Park in Taiwan. Participants included elementary school teachers and students. The results of the evaluation in this study show that the proposed EULER significantly improves student motivation and learning. Furthermore, the results of a post-study survey reveal that most student feedback is positive, further indicating the effectiveness of the EULER.

Wearing a safety helmet is important in construction and manufacturing industrial activities to avoid unpleasant situations. This safety compliance can be ensured by developing an automatic helmet detection system using various computer vision and deep learning approaches. Developing a deep-learning-based helmet detection model usually requires an enormous amount of training data. However, there are very few public safety helmet datasets available in the literature, in which most of them are not entirely labeled, and the labeled one contains fewer classes. This paper presents the Safety HELmet dataset with 5K images (SHEL5K) dataset, an enhanced version of the SHD dataset. The proposed dataset consists of six completely labeled classes (helmet, head, head with helmet, person with helmet, person without helmet, and face). The proposed dataset was tested on multiple state-of-the-art object detection models, i.e., YOLOv3 (YOLOv3, YOLOv3-tiny, and YOLOv3-SPP), YOLOv4 (YOLOv4 and YOLOv4_pacsp-x-mish), YOLOv5-P5 (YOLOv5s, YOLOv5m, and YOLOv5x), the Faster Region-based Convolutional Neural Network (Faster-RCNN) with the Inception V2 architecture, and YOLOR. The experimental results from the various models on the proposed dataset were compared and showed improvement in the mean Average Precision (mAP). The SHEL5K dataset had an advantage over other safety helmet datasets as it contains fewer images with better labels and more classes, making helmet detection more accurate.