Fall detection, particularly critical for high-risk demographics like the elderly, is a key public health concern, where timely detection can greatly minimize harm. With the advancements in radio frequency (RF) technology, radar has emerged as a powerful tool for human fall detection. Traditional machine learning (ML) algorithms, such as support vector machines (SVM) and k -nearest neighbors (kNN), have shown promising outcomes. However, deep learning (DL) approaches, notably convolutional neural networks (CNNs) and recurrent neural networks (RNNs), have outperformed in learning intricate features and managing large, unstructured datasets. This survey offers an in-depth analysis of radar-based fall detection, with emphasis on micro-Doppler, range-Doppler, and range-Doppler-angles techniques. We discuss the intricacies and challenges in fall detection and emphasize the necessity for a clear definition of falls and appropriate detection criteria, informed by diverse influencing factors. We present an overview of radar signal-processing principles and the underlying technology of radar-based fall detection, providing an accessible insight into ML and DL algorithms. After examining 74 research articles on radar-based fall detection published since 2000, we aim to bridge current research gaps and underscore the potential future research strategies, emphasizing the real-world applications possibility and the unexplored potential of DL in improving radar-based fall detection.
In this paper, we propose mmPose-FK, a novel millimeter wave (mmWave) radar-based pose estimation method that employs a dynamic forward kinematics (FK) approach to address the challenges posed by low resolution, specularity, and noise artifacts commonly associated with mmWave radars. These issues often result in unstable joint poses that vibrate over time, reducing the effectiveness of traditional pose estimation techniques. To overcome these limitations, we integrate the FK mechanism into the deep learning model and develop an end-to-end solution driven by data. Our comprehensive experiments using various matrices and benchmarks highlight the superior performance of mmPose-FK, especially when compared to our previous research methods. The proposed method provides more accurate pose estimation and ensures increased stability and consistency, which underscores the continuous improvement of our methodology, showcasing superior capabilities over its antecedents. Moreover, the model can output joint rotations and human bone lengths, which could be further utilized for various applications such as gait parameter analysis and height estimation. This makes mmPose-FK a highly promising solution for a wide range of applications in the field of human pose estimation and beyond.
5 March 2025 in IEEE Transactions on Biomedical Engineering
This study investigates a novel approach for analyzing Sit-to-Stand (STS) movements using millimeter-wave (mmWave) radar technology, aiming to develop a non-contact, privacy-preserving, and all-day operational solution for healthcare applications. A 60 GHz mmWave radar system was employed to collect radar point cloud data from 45 participants performing STS motions. Using a deep learning-based pose estimation model and Inverse Kinematics (IK), we calculated joint angles, segmented STS motions, and extracted clinically relevant features for fall risk assessment. The extracted features were compared with those obtained from Kinect and wearable sensors. While Kinect provided a reference for motion capture, we acknowledge its limitations compared to the gold-standard VICON system, which is planned for future validation. The results demonstrated that mmWave radar effectively captures general motion patterns and large joint movements (e.g., trunk), though challenges remain for more fine-grained motion analysis. This study highlights the unique advantages and limitations of mmWave radar and other sensors, emphasizing the potential of integrated sensor technologies to enhance the accuracy and reliability of motion analysis in clinical and biomedical research. Future work will expand the scope to more complex movements and incorporate high-precision motion capture systems to further validate the findings.
Unmanned aerial vehicles (UAVs) can be used as air base stations to provide fast wireless connections for ground users. Due to their constraints on both mobility and energy consumption, a key problem is how to deploy UAVs adaptively in a geographic area with changing traffic demand of mobile users, while meeting the aforemetioned constraints. In this paper, we propose an adaptive deployment strategy for UAV-aided networks based on hybrid deep reinforcement learning, where a UAV can adjust its movement direction and distance to serve users who move randomly in the target area. Through hybrid deep reinforcement learning, UAVs can be trained offline to obtain the global state information and learn a completely distributed control strategy, with which each UAV only needs to take actions based on its observed state in the real deployment to be fully adaptive. Moreover, in order to improve the speed and effect of learning, we improve hybrid reinforcement learning, by adding genetic algorithms and TD-error-based resampling optimization mechanism. Simulation results show that the hybrid deep reinforcement learning algorithm has better efficiency and robustness in multi-UAV control, and has better performance in terms of coverage, energy consumption and average throughput, by which average throughput can be increased by 20% to 60%.
Fall detection, particularly critical for high-risk demographics like the elderly, is a key public health concern where timely detection can greatly minimize harm. With the advancements in radio frequency technology, radar has emerged as a powerful tool for human detection and tracking. Traditional machine learning algorithms, such as Support Vector Machines (SVM) and k-Nearest Neighbors (kNN), have shown promising outcomes. However, deep learning approaches, notably Convolutional Neural Networks (CNN) and Recurrent Neural Networks (RNN), have outperformed in learning intricate features and managing large, unstructured datasets. This survey offers an in-depth analysis of radar-based fall detection, with emphasis on Micro-Doppler, Range-Doppler, and Range-Doppler-Angles techniques. We discuss the intricacies and challenges in fall detection and emphasize the necessity for a clear definition of falls and appropriate detection criteria, informed by diverse influencing factors. We present an overview of radar signal processing principles and the underlying technology of radar-based fall detection, providing an accessible insight into machine learning and deep learning algorithms. After examining 74 research articles on radar-based fall detection published since 2000, we aim to bridge current research gaps and underscore the potential future research strategies, emphasizing the real-world applications possibility and the unexplored potential of deep learning in improving radar-based fall detection.
1 August 2026 in IEEE Transactions on Medical Robotics and Bionics