Numerous learning-based techniques for effective human behavior identification have emerged in recent years. These techniques focus only on fundamental human activities, excluding transitional activities due to their infrequent occurrence and short period. Nevertheless, postural transitions play a critical role in implementing a system for recognizing human activity and cannot be ignored. This study aims to present a hybrid deep residual model for transitional activity recognition utilizing signal data from wearable sensors. The developed model enhances the ResNet model with hybrid Squeeze-and-Excitation (SE) residual blocks combining a Bidirectional Gated Recurrent Unit (BiGRU) to extract deep spatio-temporal features hierarchically, and to distinguish transitional activities efficiently. To evaluate recognition performance, the experiments are conducted on two public benchmark datasets (HAPT and MobiAct v2.0). The proposed hybrid approach achieved classification accuracies of 98.03% and 98.92% for the HAPT and MobiAct v2.0 datasets, respectively. Moreover, the outcomes show that the proposed method is superior to the state-of-the-art methods in terms of overall accuracy. To analyze the improvement, we have investigated the effects of combining SE modules and BiGRUs into the deep residual network. The findings indicates that the SE module is efficient in improving transitional activity recognition.

The recognition of human movements is a crucial aspect of AI-related research fields. Although methods using vision and sensors provide more valuable data, they come at the expense of inconvenience to users and social limitations including privacy issues. WiFi-based sensing methods are increasingly being used to collect data on human activity due to their ubiquity, versatility, and high performance. Channel state information (CSI), a characteristic of WiFi signals, can be employed to identify various human activities. Traditional machine learning approaches depend on manually designed features, so recent studies propose leveraging deep learning capabilities to automatically extract features from raw CSI data. This research introduces a versatile framework for recognizing human activities by utilizing CSI data and evaluates its effectiveness on different deep learning networks. A hybrid deep learning network called CNN-GRU-AttNet is proposed to automatically extract informative spatial-temporal features from raw CSI data and efficiently classify activities. The effectiveness of a hybrid model is assessed by comparing it with five conventional deep learning models (CNN, LSTM, BiLSTM, GRU, and BiGRU) on two widely recognized benchmark datasets (CSI-HAR and StanWiFi). The experimental results demonstrate that the CNN-GRU-AttNet model surpasses previous state-of-the-art techniques, leading to an average accuracy improvement of up to 4.62%. Therefore, the proposed hybrid model is suitable for identifying human actions using CSI data.

Human activity recognition (HAR) extensively uses wearable inertial sensors since this data source provides the most information for non-visual datasets’ time series. HAR research has advanced significantly in recent years due to the proliferation of wearable devices with sensors. To improve recognition performance, HAR researchers have extensively investigated other sources of biosignals, such as a photoplethysmograph (PPG), for this task. PPG sensors measure the rate at which blood flows through the body, and this rate is regulated by the heart’s pumping action, which constantly occurs throughout the body. Even though detecting body movement and gestures was not initially the primary purpose of PPG signals, we propose an innovative method for extracting relevant features from the PPG signal and use deep learning (DL) to predict physical activities. To accomplish the purpose of our study, we developed a deep residual network referred to as PPG-NeXt, designed based on convolutional operation, shortcut connections, and aggregated multi-branch transformation to efficiently identify different types of daily life activities from the raw PPG signal. The proposed model achieved more than 90% prediction F1-score from experimental results using only PPG data on the three benchmark datasets. Moreover, our results indicate that combining PPG and acceleration signals can enhance activity recognition. Although, both biosignals—electrocardiography (ECG) and PPG—can differentiate between stationary activities (such as sitting) and non-stationary activities (such as cycling and walking) with a level of success that is considered sufficient. Overall, our results propose that combining features from the ECG signal can be helpful in situations where pure tri-axial acceleration (3D-ACC) models have trouble differentiating between activities with relative motion (e.g., walking, stair climbing) but significant differences in their heart rate signatures.

Falls are a serious problem that are often experienced by the elderly in performing activities in daily living. In recent years, the use of smartphone sensors in fall detection and activity daily living (ADL) classification has been studied to explore and understand human behaviours. In this paper, we investigate the role of the multi-sensors in fall detection and the ADL classification problem. We present ensemble learning-based approaches to improve recognition performance. We evaluate their roles on two body positions, which are the arm position and the waist position, while recognizing six ADL activities: standing, sitting, laying, walking, walking upstairs, and walking downstairs, and two different falls: falling after walking and falling after standing. From the experimental results, the ensemble learning-based approaches can improve the recognition performance for using only accelerometer data at the arm position with an accuracy of 94.8750 percent. Moreover, at the waist position, the ensemble approaches can improve the recognition performance for using both accelerometer data and gyroscope data with an accuracy of 100.00 percent.

With the fast expansion of digital technologies and sporting events, interpreting sports data has become an immensely complicated endeavor. Internet-sourced sports big data exhibit a significant development trend. Big data in sports offer a wealth of information on sportspeople, coaching, athletics, swimming, and badminton. Today, various sports data are freely accessible, and incredible data analysis tools based on wearable sensors have been established, allowing us to investigate the usefulness of these data thoroughly. In this research, we investigate the detection of badminton action and player evaluation based on movement data captured by wearable sensors. Movement data captured by an accelerometer, gyroscope, and magnetometer are utilized for training and validating a classification model for badminton actions. In addition, the movement signals are used to train a player evaluation model employing a deep residual network. To assess our suggested technique, we utilized a publicly available benchmark dataset consisting of inertial measurement unit (IMU) sensors attached to every investigator’s dominant wrist, palm, and both legs. The experimental findings indicate that the proposed deep residual network obtained good performance with a maximum accuracy of 98.00% for identifying badminton activities and 98.56% for evaluating badminton players.

Thailand, like other countries worldwide, has experienced instability in recent years. If current trends continue, the number of crimes endangering people or property will expand. Closed-circuit television (CCTV) technology is now commonly utilized for surveillance and monitoring to ensure people’s safety. A weapon detection system can help police officers with limited staff minimize their workload through on-screen surveillance. Since CCTV footage captures the entire incident scenario, weapon detection becomes challenging due to the small weapon objects in the footage. Due to public datasets providing inadequate information on our interested scope of CCTV image’s weapon detection, an Armed CCTV Footage (ACF) dataset, the self-collected mockup CCTV footage of pedestrians armed with pistols and knives, was collected for different scenarios. This study aimed to present an image tilling-based deep learning for small weapon object detection. The experiments were conducted on a public benchmark dataset (Mock Attack) to evaluate the detection performance. The proposed tilling approach achieved a significantly better mAP of 10.22 times. The image tiling approach was used to train different object detection models to analyze the improvement. On SSD MobileNet V2, the tiling ACF Dataset achieved an mAP of 0.758 on the pistol and knife evaluation. The proposed method for enhancing small weapon detection by using the tiling approach with our ACF Dataset can significantly enhance the performance of weapon detection.