Dataset Hub

The dataset is for fall detection using wearable devices, containing heart‑rate sensor and accelerometer data. Collected from 21 participants across 19 different scenarios, including 6 fall types, 9 daily activities, and 4 near‑fall situations. Detailed information includes sensor attachment location, sampling rate, scenario description, and participant demographics.

MPFDD is a multi‑person fall dataset comprising indoor and outdoor scenes with 2 to 5 participants. The dataset contains 220 videos, of which 80 depict daily activities and 140 depict falls. Indoor scenes feature an activity room equipped with chairs, tables, computers, etc., while outdoor scenes are open areas primarily with tables and chairs. The dataset details video distribution across different participant counts.