Pathological Gait Datasets Overview

Dataset Content

• Type: skeletal dataset covering six gait types: normal, painful, stiff‑leg, swaying, stepping, and Trendelenburg.
• Volume: 10 subjects × 6 gait types × 120 instances.

Data Format

• Structure: timestamps followed by sequences of joint coordinates (x, y, z).
• Example: time, 0, joint0_x, joint0_y, joint0_z, 1, joint1_x, joint1_y, joint1_z, 2, joint2_x, joint2_y, joint2_z, ...

Data Collection

• Devices: six Kinect v2 systems.
• Calibration: all sensors were calibrated using ArUco markers to align coordinate systems.

Reference

• K. Jun, Y. Lee, S. Lee, D. Lee and M. S. Kim, "Pathological Gait Classification Using Kinect v2 and Gated Recurrent Neural Networks," IEEE Access, vol. 8, pp. 139881‑139891, 2020, doi: 10.1109/ACCESS.2020.3013029.

Contact

• Email: kooksung930@gm.gist.ac.kr