Multimodal gait dataset with floor accelerations, inertial sensor data, and synchronized video from controlled...

Area of Science:

• Biomechanics
• Structural Engineering
• Data Science

Background:

• Human locomotion generates vibrations that interact with structures.
• Accurate gait analysis requires multimodal data for comprehensive understanding.
• Existing datasets may lack synchronized measurements of gait and resulting vibrations.

Purpose of the Study:

• To provide a multimodal dataset for gait analysis and human-structure interaction research.
• To enable the development and validation of models for floor vibration and locomotion.
• To facilitate data-driven movement analysis and machine learning applications.

Main Methods:

• Controlled walking trials with a single participant on a 10m walkway.
• Collection of floor acceleration signals using seismic accelerometers.
• Acquisition of body-mounted inertial measurements using OPAL-APDM sensors.
• Synchronized video recordings of all trials.

Main Results:

• A dataset comprising floor vibrations, inertial sensor data, and synchronized videos was generated.
• Data captured walking trials at controlled cadences (90, 120, 130 bpm).
• The dataset is publicly available for research purposes.

Conclusions:

• The dataset is valuable for advancing research in gait analysis and structural dynamics.
• It supports the creation and testing of computational models for human-structure interaction.
• Facilitates the development of machine learning algorithms for analyzing human movement.