Record movement in context
The project includes activities such as walking, running, and using stairs.
IEEE ICRA · 2025 / PERCEPTION & CONTROL
HelmetPoser.
Understanding motion from a sensor you can wear.
A helmet-mounted IMU dataset for estimating human head motion across activities and environments.
- Inertial sensing
- Human motion
- Dataset
01 / OVERVIEW
The idea.
HelmetPoser: A Helmet-Mounted IMU Dataset for Data-Driven Estimation of Human Head Motion in Diverse Conditions
IEEE International Conference on Robotics and Automation (ICRA) · 2025
HelmetPoser studies data-driven human head-motion estimation using a helmet-mounted inertial sensor. The dataset brings together different activities and recording conditions to support evaluation beyond a single controlled motion.
- Helmet-mounted IMU recordings
- Diverse movement conditions
- Data-driven motion estimation
RESEARCH VISUALS
Watch the experiment.
Walking activity preview from the official HelmetPoser project page. The original video introduces the full dataset.
02 / IMPLEMENTATION
How it works.
- Wear
Record motion with the instrumented helmet.
- Move
Capture different activities and conditions.
- Align
Associate inertial signals with reference motion.
- Evaluate
Study data-driven motion estimation.
Support motion-estimation research
The recordings enable experiments with helmet-mounted inertial measurements.
Explore the dataset
The authors’ project page provides the dataset overview and demonstrations.
03 / RESOURCES
Read further.
The paper, original demonstrations, and available project materials.