Abstract:Gesture recognition, being a natural and convenient means of human-computer interaction, has emerged as a prominent area of research. The goal of accurate and swift gesture recognition has driven numerous studies, as it plays a crucial role in meeting practical application demands. Real-time monitoring of surface electromyography (sEMG) signals collected from sEMG sensor during gesture execution enables the extraction of rich gesture action features, prompting researchers to delve into sEMG-based gesture recognition algorithms. However, existing algorithms often struggle to strike a favorable balance between prediction accuracy and computational efficiency. In response to these challenges, this paper proposes a new gesture recognition model leveraging multi-channel sEMG signals captured by a sEMG sensor. The approach involves employing a lightweight convolutional neural network (CNN) model as the backbone network, supplemented by a lightweight feature extraction block (LFE) to effectively aggregate multi-scale features. Additionally, the model benefits from pre-training weights based on transfer learning, enhancing feature extraction capabilities and generalization ability for faster convergence of the prediction model. Moreover, the introduction of a plug-and-play block called the broad attention learning block (BAL) significantly improves the overall recognition performance of the proposed model. To better capture contextual information from multi-channel sEMG signals, a pyramid input with varying resolutions is constructed, elevating the model's multi-scale feature extraction capability. The proposed model is evaluated and validated on a dynamic gesture dataset collected using wearable sensor (Myo). A series of ablation experiments, comparative analyses, and visualization outcomes unequivocally showcase the remarkable efficacy and superiority of our proposed method. Comparisons with other advanced models reveal that the proposed model successfully achieves a balanced compromise between prediction accuracy and processing speed.

Lightweight Online Semi-Supervised Learning Algorithm for Ultrasonic Gesture Recognition.

Lightweight Online Semisupervised Learning for Ultrasonic Radar-Based Dynamic Hand Gesture Recognition

Efficiently User-Independent Ultrasonic-Based Gesture Recognition Algorithm

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

Semi-Supervised FMCW Radar Hand Gesture Recognition via Pseudo-Label Consistency Learning

Efficient High Cross-User Recognition Rate Ultrasonic Hand Gesture Recognition System

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

Semg-Based Inter-Session Hand Gesture Recognition Via Domain Adaptation with Locality Preserving and Maximum Margin

A Lightweight Hand-Gesture Recognition Network With Feature Fusion Prefiltering and FMCW Radar Spatial Angle Estimation

Improved Network and Training Scheme for Cross-Trial Surface Electromyography (Semg)-Based Gesture Recognition.

mmGesture: Semi-supervised gesture recognition system using mmWave radar

A wearable sensor-based dynamic gesture recognition model via broad attention learning

sEMG-based Fine-grained Gesture Recognition via Improved LightGBM Model

Unsupervised Domain Adaptation for RF-based Gesture Recognition

Bayesian Filter-Based Ultrasonic Continuous Hand Gesture Recognition for Low-Resource Platform

Online cross session electromyographic hand gesture recognition using deep learning and transfer learning

Boosting Semi-Supervised Face Recognition With Noise Robustness

Unsupervised Domain Adaptation for Device-free Gesture Recognition

One-Shot-Learning Gesture Segmentation and Recognition Using Frame-Based PDV Features.

Simple and Efficient Gesture Recognition Based on Frequency-Modulated Continuous Wave Radar

A Novel Dual-Model Adaptive Continuous Learning Strategy for Wrist-sEMG Real-Time Gesture Recognition