Abstract:In this paper, we present a putEMG dataset intended for evaluation of hand gesture recognition methods based on sEMG signal. The dataset was acquired for 44 able-bodied subjects and include 8 gestures (3 full hand gestures, 4 pinches, and idle). It consists of uninterrupted recordings of 24 sEMG channels from the subject's forearm, RGB video stream and depth camera images used for hand motion tracking. Moreover, exemplary processing scripts are also published. putEMG dataset is available under Creative Commons Attribution-NonCommercial 4.0 International (CC BY-NC 4.0) license at: <a class="link-external link-https" href="https://www.biolab.put.poznan.pl/putemg-dataset/" rel="external noopener nofollow">this https URL</a>. The dataset was validated regarding sEMG amplitudes and gesture recognition performance. The classification was performed using state-of-the-art classifiers and feature sets. Accuracy of 90% was achieved for SVM classifier utilising RMS feature and for LDA classifier using Hudgin's and Du's feature sets. Analysis of performance for particular gestures showed that LDA/Du combination has significantly higher accuracy for full hand gestures, while SVM/RMS performs better for pinch gestures. Presented dataset can be used as a benchmark for various classification methods, evaluation of electrode localisation concepts, or development of classification methods invariant to user-specific features or electrode displacement.

What problem does this paper attempt to address?

The problem that this paper attempts to solve is to develop a reliable gesture recognition dataset for evaluating gesture recognition methods based on surface electromyography (sEMG) signals. Specifically, several key issues mentioned in the article include: 1. **Electrode Design and User Comfort**: Develop electrodes that can ensure a constant input impedance while providing sufficient user comfort and ease of use, especially for long - term use outside the laboratory. 2. **Gesture Recognition Accuracy**: Achieve high - precision gesture recognition and maintain high accuracy even among different users. 3. **Simplify the Calibration Process**: Reduce complex calibration steps to lower the entry barrier for users and prevent users from being deterred by the cumbersome calibration process. 4. **Individual Differences and Signal Non - stationarity**: Cope with the high individualization and non - stationarity of electromyography signals, which makes it difficult to develop a universal gesture recognition system for different users. 5. **Influence of External Factors**: Consider the influence of factors such as external mechanical loads on electromyography signals, which may cause changes in signal characteristics. 6. **Influence of Electrode Position Changes**: Since the electrode position may change when the user repeatedly wears the device, it is necessary to develop a gesture recognition method that is not sensitive to the electrode position. To solve the above problems, the authors proposed the putEMG dataset, which is a gesture recognition dataset containing surface electromyography signals, RGB video streams, and depth - camera images of 44 healthy subjects. This dataset covers 8 gestures (3 whole - hand gestures, 4 pinch - grip gestures, and one idle state), and each gesture has 20 repeated recordings. The characteristics of the dataset include: - **A Large Number of Repetitions**: Each gesture has 20 repetitions, which improves the robustness of the data. - **Multi - session Recordings**: Each subject conducted two experiments with an interval of one week to consider long - term changes. - **Sparse Electrode Configuration**: Recordings were made using 24 electrodes (3 elastic bands with 8 electrodes each), which is closer to the electrode configuration in practical applications. - **Synchronous Gesture Tracking**: Use depth sensors and high - definition cameras for gesture tracking to ensure the authenticity and accuracy of gesture execution. Through this dataset, researchers can develop and test gesture recognition classification methods that are not affected by user - specific characteristics or electrode displacement, and evaluate the effectiveness of the electrode positioning concept. In addition, the dataset also provides example processing scripts to facilitate researchers to quickly adopt and process data.

putEMG -- a surface electromyography hand gesture recognition dataset

Dataset for multi-channel surface electromyography (sEMG) signals of hand gestures

Gesture recognition by instantaneous surface EMG images

FORS-EMG: A Novel sEMG Dataset for Hand Gesture Recognition Across Multiple Forearm Orientations

Surface EMG-Based Inter-Session Gesture Recognition Enhanced by Deep Domain Adaptation.

sEMG-based Gesture-Free Hand Intention Recognition: System, Dataset, Toolbox, and Benchmark Results

Effective features extraction and selection for hand gesture recognition using sEMG signal

emg2pose: A Large and Diverse Benchmark for Surface Electromyographic Hand Pose Estimation

Open Access Dataset, Toolbox and Benchmark Processing Results of High-Density Surface Electromyogram Recordings

Gesture Recognition by Ensemble Extreme Learning Machine Based on Surface Electromyography Signals

Data Augmentation of Surface Electromyography for Hand Gesture Recognition

Semi-Supervised Learning for Surface EMG-based Gesture Recognition.

Semg-Based Gesture Recognition with Embedded Virtual Hand Poses and Adversarial Learning.

Forearm sEMG data from young healthy humans during the execution of hand movements

Machine Learning-based sEMG Signal Classification for Hand Gesture Recognition

SeNic: An Open Source Dataset for sEMG-Based Gesture Recognition in Non-Ideal Conditions

Finger Movement Recognition via High-Density Electromyography of Intrinsic and Extrinsic Hand Muscles

EMG subspace alignment and visualization for cross-subject hand gesture classification

EMG-Based Hand Gesture Recognition through Diverse Domain Feature Enhancement and Machine Learning-Based Approach

A robust gesture recognition algorithm based on surface EMG

Enhanced Hand Gesture Recognition with Surface Electromyogram and Machine Learning