Upper_Arm_sEMG_Movement_Classification

Evaluation of the Feasibility of Classifying Various Upper Limb Movements Using Multichannel Upper-Arm sEMG Signals

Daehui Kim¹, Junghun Kim²*, Sang-Il Choi²*

¹ Department of Computer Software, Daegu Catholic University, Gyeongsangbuk-do, 38430, Republic of Korea
² School of Computer Software, Daegu Catholic University, Gyeongsangbuk-do, 38430, Republic of Korea

* Corresponding authors: fainal2@cu.ac.kr; sangilchoi@cu.ac.kr

---

Overview

This repository provides surface electromyography (sEMG) data recorded from upper-arm muscles during various upper limb movements.

The dataset was constructed to evaluate the feasibility of classifying elbow, forearm, wrist, and hand movements using only multichannel upper-arm sEMG signals.

---

Dataset Description

sEMG signals were recorded from five upper-arm muscle sites.

Channel	Muscle
CH1	Biceps brachii long head
CH2	Biceps brachii short head
CH3	Triceps brachii long head
CH4	Triceps brachii lateral head
CH5	Brachialis

The signals were collected at a sampling frequency of 1000 Hz.

The experiment was conducted under two posture conditions.

• Static condition: elbow fixed on a desk
• Free condition: elbow not fixed

The following eight upper limb movements were performed.

Label	Movement
CH	Hand Close
OH	Hand Open
WF	Wrist Flexion
WE	Wrist Extension
PRO	Forearm Pronation
SUP	Forearm Supination
EF	Elbow Flexion
EE	Elbow Extension

---

Repository Structure

Hand_Gesture_Classification/
├── data/
│   ├── CH1_a.csv
│   ├── CH1_b.csv
│   ├── CH1_c.csv
│   ├── ...
│   └── ...
├── LICENSE
└── README.md

---

File Naming Rule

The data files are named according to the following format.

[movement label][trial number]_[participant label].csv
[movement label][trial number]_s_[participant label].csv

Examples:

CH1_a.csv
CH1_s_a.csv
CH2_a.csv
CH2_s_a.csv

The file name consists of the movement label, trial number, posture condition, and participant label.

• CH, OH, WF, WE, PRO, SUP, EF, and EE indicate the movement labels.
• The number after the movement label indicates the trial session number.
• Files without _s_ indicate data collected under the static condition.
• Files containing _s_ indicate data collected under the free condition.
• The final alphabetic label indicates the participant.

For example:

• CH1_a.csv indicates data from participant a, collected during the 1st trial session of the hand close movement under the static condition.
• CH1_s_a.csv indicates data from participant a, collected during the 1st trial session of the hand close movement under the free condition.

Each movement was recorded for three trial sessions.

---

Data Usage

This dataset can be used for research on:

• sEMG-based upper limb movement classification
• hand and wrist gesture recognition
• upper-arm sEMG signal analysis
• machine learning-based biosignal classification
• comparison of sEMG classification performance under different posture conditions

---

Ethics Statement

This dataset was collected after Institutional Review Board (IRB) approval and written informed consent from all participants.

---

Citation

If you use this dataset or repository in your research, please cite the related paper.

---

License

This dataset is provided for research and educational purposes only.

Please see the LICENSE file for details.

---

Contact

For questions regarding this repository, please contact the email below.

eogml7732@gmail.com