Datasets
Standard Dataset
Gait analysis
- Citation Author(s):
- Submitted by:
- Fo Hu
- Last updated:
- Sat, 10/19/2024 - 04:17
- DOI:
- 10.21227/6xn4-pv39
- Data Format:
- License:
90 Views- Categories:
- Keywords:
Abstract
This custom dataset was created to support gait recognition research using Inertial Measurement Units (IMUs), which capture acceleration, angular velocity, and orientation data from key body locations (e.g., ankles, waist, wrists). It includes recordings from [insert number] participants performing various walking tasks under different conditions, such as normal and fast walking or navigating obstacles. The dataset provides time-series data suitable for both traditional feature-based analysis and deep learning approaches. It aims to facilitate research in biometric authentication, activity recognition, abnormality detection, and wearable healthcare applications, offering a flexible resource for developing real-world gait recognition systems.
This dataset is organized by participant IDs, with each trial containing time-series data (e.g., acceleration, angular velocity) captured under various walking tasks. Data files are provided in CSV/XLSX format, accompanied by a metadata file with participant details and task descriptions. Preprocessing steps such as synchronization, filtering, and segmentation are recommended for optimal use. The dataset supports both traditional feature-based methods and deep learning approaches for applications like biometric authentication, activity recognition, and healthcare monitoring. Please cite [insert citation] if you use this dataset in your research.
Comments
It is worth recalling that all measures (except Sl) range from 0 to 1, what can be directly understood by physicians. This fact makes them semantic, easy-to-interpret features.