Heart Rate & IMU sensor data for fall detection (HIFD dataset)

Dataset Overview

Basic Information

• Dataset Name: Heart Rate & IMU sensor data for fall detection (HIFD dataset)
• Sensor Position: Left wrist
• Number of Subjects: 21 (13 males, 8 females)
• Sampling Rate: 50 Hz
• Number of Scenarios: 19 (including 6 fall types, 9 daily activities, 4 near‑fall situations)
• Accelerometer Range: ±16 g
• Heart‑Rate Signal Processing: Provides RR intervals for easy extraction

Scenario Description

• Fall Scenarios:
  • Clockwise forward fall
  • Clockwise backward fall
  • Right‑to‑left side fall
  • Counter‑clockwise forward fall
  • Counter‑clockwise backward fall
  • Left‑to‑right side fall
• Daily Activity Scenarios:
  • Lying down and getting up on bed
  • Sitting down and standing up on a chair
  • Clapping hands
  • Dressing
  • Eating
  • Brushing hair
  • Tying shoes
  • Going up and down stairs
  • Brushing teeth
  • Walking
  • Washing hands
  • Writing
  • Fast zip‑up and zip‑down

Data Representation

• Accelerometer Signals:
  • ax: Acceleration on X axis (g)
  • ay: Acceleration on Y axis (g)
  • az: Acceleration on Z axis (g)
• Gyroscope Signals:
  • w, x, y, z: Quaternion from gyroscope
  • droll, dpitch, dyaw: Angular velocity from gyroscope
• Heart‑Rate Signal:
  • heart: PPG sensor signal
• Time:
  • time: Real‑time timestamp

Subject Information

• Age & Gender:
  • Age range: 21‑32 years
  • Gender distribution: 13 males, 8 females

The above information summarises the basic situation of the HIFD dataset, including sensor configuration, subject characteristics, data‑collection scenarios and specific data representation methods.