An IMU-based dataset of falls, activities of daily living, and prayer movements (AybuFall).

objectivesPublicly available datasets are essential for the development, evaluation, and benchmarking of fall detection and human activity recognition algorithms. Although numerous datasets include falls and activities of daily living (ADLs), prayer movements-despite exhibiting motion patterns that may resemble falls-remain largely underrepresented. The objective of this study is to present a publicly available IMU-based dataset that explicitly includes prayer movements alongside falls and ADLs, thereby addressing an important gap in existing datasets and supporting methodological research on activity classification and false-positive reduction. DATA DESCRIPTION: The dataset comprises motion recordings of 11 types of fall movements, 13 types of activities of daily living (ADLs), and 5 types of prayer movements. Data were collected from 17 healthy young adult participants using two wearable IMU sensors placed on the forehead and forearm. Each activity was performed three times by each participant. Tri-axial accelerometer, gyroscope, and magnetometer signals were recorded at a sampling frequency of 200 Hz. All recordings were manually labeled by direct observation during data acquisition. The dataset is publicly available and systematically organized to support algorithm development, benchmarking, and reproducible research in fall detection and human activity recognition. Although data were collected from young adults, the dataset is intended as a controlled reference resource, and applicability to other populations requires further validation.