Article
Engineering, Electrical & Electronic
Myung-Kyu Yi, Wai-Kong Lee, Seong Oun Hwang
Summary: Human Activity Recognition (HAR) is an important part of human life care. Through rigorous analysis of various HAR datasets, we propose a lightweight approach using statistical feature extraction to discriminate between static and dynamic activities. By replacing the first-level ML classifier with this technique, we achieve higher accuracy with less computational and memory consumption. The proposed HAR method, combined with Random Forest and Convolutional Neural Networks, achieves state-of-the-art results and is practical for wearable devices using a single accelerometer.
IEEE TRANSACTIONS ON CONSUMER ELECTRONICS
(2023)
Review
Engineering, Electrical & Electronic
E. Ramanujam, Thinagaran Perumal, S. Padmavathi
Summary: Human Activity Recognition (HAR) is the field of inferring human activities from signals acquired through sensors of smartphones and wearable devices, mainly for smart home and elderly care. Current techniques mostly use Deep Learning for feature extraction and classification efficiency, but there are challenges and issues that require future research and improvements.
IEEE SENSORS JOURNAL
(2021)
Article
Engineering, Electrical & Electronic
Mahdi Pedram, Ramesh Kumar Sah, Seyed Ali Rokni, Marjan Nourollahi, Hassan Ghasemzadeh
Summary: Advances in embedded systems have led to the integration of wearable sensors in health monitoring. However, due to the personalized nature of human movement and the limitations of embedded sensors, a resource-efficient framework is needed for real-time activity recognition.
IEEE SENSORS JOURNAL
(2022)
Article
Engineering, Electrical & Electronic
Zhixin Pan, Huihui Chen, Weizhao Zhong, Aiguo Wang, Chundi Zheng
Summary: This article proposes a CNN-based behavioral recognition method for automatic monitoring of lactating sows. The behavior data streams are collected through wearable sensors and used to recognize six types of behaviors. By leveraging CNNs and using data augmentation, the proposed method achieves higher accuracy in distinguishing static behaviors compared to traditional machine learning methods. The research results have important implications for behavioral monitoring and health assessment of lactating sows.
IEEE SENSORS JOURNAL
(2023)
Article
Engineering, Electrical & Electronic
Ying Li, Junsheng Wu, Weigang Li, Aiqing Fang, Wei Dong
Summary: The sensor-based human activity recognition (SHAR) task aims to recognize signals collected by sensors in intelligent devices to assist people in their daily lives. Deep learning is being studied for combining with SHAR. To address the challenge of maintaining efficiency, an effective sensor signal representation method, called the temporal-spatial dynamic convolutional network, is presented. Extensive experiments demonstrate the superiority of this method over deep learning baselines and existing SHAR works on benchmark SHAR datasets.
IEEE TRANSACTIONS ON INSTRUMENTATION AND MEASUREMENT
(2023)
Article
Engineering, Electrical & Electronic
Yves Luduvico Coelho, Francisco de Assis Souza dos Santos, Anselmo Frizera-Neto, Teodiano Freire Bastos-Filho
Summary: Human Activity Recognition (HAR) has gained increasing attention from researchers and industry, requiring the design of small, lightweight, powerful, and low-cost smart sensors for practical HAR systems on wearable devices. Edge computing presents an energy-efficient solution that offers real-time response and privacy requirements for HAR applications.
IEEE SENSORS JOURNAL
(2021)
Article
Chemistry, Analytical
Johannes Link, Timur Perst, Maike Stoeve, Bjoern M. Eskofier
Summary: In this study, the researchers recorded and annotated IMU data of different types of Ultimate Frisbee throws. They used Convolutional Neural Networks (CNNs) and transfer learning to classify and recognize the actions. The proposed pipeline achieved high accuracy, especially in distinguishing fine-grained classes. The study also compared the results to a transfer learning-based approach using a different sports dataset and analyzed the impact of transfer learning on a reduced dataset without data augmentations.
Article
Computer Science, Artificial Intelligence
Uriel Martinez-Hernandez, Mohammed I. Awad, Abbas A. Dehghani-Sanij
Summary: This study introduces a novel learning architecture for the recognition and prediction of walking activity and gait period using wearable sensors. The architecture combines CNN and PIG methods for recognition and prediction, and achieves high accuracy through adaptive combination. Experimental results demonstrate the effectiveness of this approach in accurate recognition and prediction of walking activity and gait period.
Article
Computer Science, Artificial Intelligence
Cagatay Berke Erdas, Selda Guney
Summary: The use of wearable sensors for activity recognition tasks has become widespread, with deep learning algorithms such as Convolutional Neural Networks proving to be effective. In this study, data from accelerometer sensors was used to feed deep learning algorithms, forming consecutive raw samples of the same activity to capture patterns and preserve the continuous structure of movement.
NEURAL PROCESSING LETTERS
(2021)
Article
Computer Science, Artificial Intelligence
Quansheng Xu, Xifei Wei, Ruxue Bai, Shiming Li, Zong Meng
Summary: This paper proposes a fast and robust hybrid model to address the transfer issues in wearable sensor based human activity recognition. The model utilizes shared features and domain adaptation methods to quickly adapt to new sensor positions and subjects with only a few annotated data, significantly improving the accuracy compared to traditional methods.
EXPERT SYSTEMS WITH APPLICATIONS
(2023)
Article
Chemistry, Analytical
Luis Sigcha, Ignacio Pavon, Nelson Costa, Susana Costa, Miguel Gago, Pedro Arezes, Juan Manuel Lopez, Guillermo De Arcas
Summary: A novel approach using smartwatches and multitask classification models can accurately assess resting tremor in Parkinson's disease patients without interfering with their daily lives. Results show high agreement with clinical assessments, indicating the potential for early-stage disease monitoring and improvement of Parkinson's disease clinical evaluation.
Article
Biochemistry & Molecular Biology
Ummara Ayman, Muhammad Sultan Zia, Ofonime Dominic Okon, Najam-ur Rehman, Talha Meraj, Adham E. Ragab, Hafiz Tayyab Rauf
Summary: In this research, a Deep Learning model was proposed to detect epileptic seizures automatically. By extracting features and classifying EEG data, the model achieved high accuracy and AUC values. The performance of the proposed model was evaluated and compared with other models using various performance parameters.
Article
Chemistry, Analytical
Mohamed Elshafei, Diego Elias Costa, Emad Shihab
Summary: This research investigates the impact of muscle fatigue on Human Activity Recognition (HAR) systems, using biceps concentration curls as an example. Findings show that fatigue prolongs completion time of later sets and decreases muscular endurance, leading to changes in data patterns and affecting the performance of subject-specific and cross-subject models. Feedforward Neural Network (FNN) exhibits the best performance in both types of models.
Article
Chemistry, Analytical
Mohamed Elshafei, Diego Elias Costa, Emad Shihab
Summary: Wearables-based HAR systems are effective for athlete performance monitoring, but user data variability can affect their performance. This study introduces a personalized model that outperforms cross-subject and subject-specific models by sourcing similar data from the crowd.
Article
Computer Science, Artificial Intelligence
Zijia Zhang, Yaoming Cai, Wenyin Gong
Summary: This paper presents a novel semi-supervised learning framework, Graph Convolutional Extreme Learning Machines (GCELM), for handling graph data in non-Euclidean domains. The proposed methods achieve significantly better results than previous methods on 36 benchmark datasets, thanks to the use of random graph convolution and a voting ensemble strategy.
EXPERT SYSTEMS WITH APPLICATIONS
(2023)
