gait-analysis-dataset

Dataset Overview

Dataset Name

• gait-analysis-dataset

Data Origin

• The dataset originates from two studies:
  • "Signal Processing and Machine Learning for Diplegia Classification"
  • "Gait-Based Diplegia Classification Using LSMT Networks"

Dataset Content

• The dataset includes 1,139 trials from 178 patients with cerebral palsy at various stages.
• Data were collected using high‑frequency VICON cameras at an Italian hospital.
• Warning: Some trials may be invalid (e.g., markers invalid in a sequence).

Data Structure

• Each .npy file contains a variable number of frames.
• Each frame records 3D coordinates of 19 markers (the first three elements) and a validation flag.
• File path structure: base_folder/class_label/subject_label/.npy

Dataset Access

• The dataset can be obtained at this link: HERE

Citation Information

• If used in research, please cite the following two papers:
  • Bergamini, Luca et al. "Signal Processing and Machine Learning for Diplegia Classification." International Conference on Image Analysis and Processing (2017): 97‑108.
  • Ferrari, Alberto et al. "Gait-Based Diplegia Classification Using LSMT Networks." Journal of Healthcare Engineering 2019.