Fall Detection IMU Dataset for Wearable Applications

This dataset contains synchronized tri-axial accelerometer and gyroscope recordings collected using a single neck-mounted IMU (LSM6DS3) during controlled trials of daily life activities and simulated falls. Approximately 1500 motion trials across 15 activity categories were recorded, including walking, sitting, bending, standing, and different fall events (forward, backward, lateral). Data are stored in CSV format, where each row corresponds to a timestamped sensor sample with acceleration (X, Y, Z), gyroscope (X, Y, Z), and associated activity labels. The dataset has been segmented into fixed-size windows for machine learning model development. This resource supports research in fall detection, human activity recognition, wearable sensing, and TinyML-based edge deployment.