A Synchronized Multimodal Dataset of Simulated Fall Events: mmWave Radar with Markerless Motion Capture Validation

This experiment presents a novel dataset acquired using a mmWave radar sensor and a reference markerless optical system (Theia Markerless) in the LABIOMEP laboratory in Porto, Portugal. The dataset consists of 40 simulated fall events (syncope and tripping), divided into fast and slow types, with each fall recorded synchronously by both the radar and the markerless camera system tracking the body’s center of mass. The aim is to provide the research community with high-quality reference data for training and validating fall detection algorithms and for further studies in biomechanics and assistive technology.