Francesca Palermo

Francesca Palermo

NinaPro DB6 Dataset

The NinaPro DB6 dataset is part of the NinaPro initiative, supporting research in surface electromyography (sEMG), motor intention decoding, and prosthetic control.

DB6 focuses on the repeatability of hand movement classification. It contains multi-session sEMG recordings collected from 10 subjects performing a set of functional grasps across multiple days and sessions. This structure allows researchers to explore inter-session variability, domain adaptation, and the robustness of machine learning models applied to sEMG signals.

I directly contributed to the acquisition of DB6 (100 total acquisitions), which is openly available to the research community.

Information

The dataset includes:

  • 10 subjects, each completing multiple daily sessions
  • 7 grasp types, repeated 12 times, twice per day, across 5 days
  • High-quality sEMG recordings suitable for:
    • Pattern recognition
    • Repeatability analysis
    • Domain adaptation
    • Robust cross-session training
    • Benchmarking sEMG-based machine learning algorithms

DB6 documentation and instructions:
🔗 https://ninapro.hevs.ch/instructions/DB6.html

Data Structure

  • Raw and processed sEMG recordings
  • Labels describing hand movements
  • Session identifiers for longitudinal analysis
  • Metadata for subject and session tracking

The dataset is suitable for supervised learning, time-series modelling, deep learning, and robustness evaluation of prosthetic control systems.

👉 Official DB6 dataset:
https://ninapro.hevs.ch/instructions/DB6.html

Info and Queries

For enquiries related to this dataset or associated research, please contact
Francesca Palermo – palermo.francesca21@gmail.com

Acknowledgments

If you use this dataset, please cite the following paper, which introduced DB6 and analysed repeatability:

Palermo, F., Cognolato, M., Gijsberts, A., Müller, H., Caputo, B., Atzori, M.
Repeatability of grasp recognition for robotic hand prosthesis control based on sEMG data.
In 2017 International Conference on Rehabilitation Robotics (ICORR), pp. 1154–1159. IEEE.