Hand gestures classification of sEMG signals based on BiLSTM- Metaheuristic Optimization and Hybrid U-Net-MobileNetV2 Encoder Architecture DOI Creative Commons

Safoura Farsi Khavari,

Khosro Rezaee, Mojtaba Ansari

et al.

Research Square (Research Square), Journal Year: 2024, Volume and Issue: unknown

Published: April 17, 2024

Abstract Surface electromyography (sEMG) data has been extensively utilized in deep learning algorithms for hand movement classification. This paper aims to introduce a novel method gesture classification using sEMG data, addressing accuracy challenges seen previous studies. We propose U-Net architecture incorporating MobileNetV2 encoder, enhanced by Bidirectional Long Short-Term Memory (BiLSTM) and metaheuristic optimization spatial feature extraction motion recognition. Bayesian is employed as the approach optimize BiLSTM model's architecture. To tackle non-stationarity of signals, we utilize windowing overlapping method, augmented with additional signals architectures. The encoder extract relevant features from spectrogram images. Edge computing integration leveraged further enhance innovation enabling real-time processing decision-making closer source. Six standard databases were utilized, achieving an average 90.23% our proposed model, showcasing 3-4% improvement 10% variance reduction. Notably, Mendeley Data, BioPatRec DB3, DB1 surpassed advanced models their respective domains accuracies 88.71%, 90.2%, 88.6%, respectively. Experimental results underscore significant enhancement generalizability recognition robustness. offers fresh perspective on prosthetic management human-machine interaction, emphasizing its efficacy improving reducing control interaction machines through edge integration.

Language: Английский

Neuromorphic hardware for somatosensory neuroprostheses DOI Creative Commons
Elisa Donati, Giacomo Valle

Nature Communications, Journal Year: 2024, Volume and Issue: 15(1)

Published: Jan. 16, 2024

Abstract In individuals with sensory-motor impairments, missing limb functions can be restored using neuroprosthetic devices that directly interface the nervous system. However, restoring natural tactile experience through electrical neural stimulation requires complex encoding strategies. Indeed, they are presently limited in effectively conveying or sensations by bandwidth constraints. Neuromorphic technology, which mimics behavior of neurons and synapses, holds promise for replicating touch, potentially informing neurostimulation design. this perspective, we propose incorporating neuromorphic technologies into neuroprostheses could an effective approach developing more human-machine interfaces, leading to advancements device performance, acceptability, embeddability. We also highlight ongoing challenges required actions facilitate future integration these advanced technologies.

Language: Английский

Citations

35
Edge AI: A Taxonomy, Systematic Review and Future Directions DOI
Sukhpal Singh Gill, Muhammed Golec,

Jianmin Hu

et al.

Cluster Computing, Journal Year: 2024, Volume and Issue: 28(1)

Published: Oct. 18, 2024

Language: Английский

Citations

18
On-Device Deep Learning for Mobile and Wearable Sensing Applications: A Review DOI
Özlem Durmaz İncel, Sevda Özge Bursa

IEEE Sensors Journal, Journal Year: 2023, Volume and Issue: 23(6), P. 5501 - 5512

Published: Feb. 7, 2023

Although running deep-learning (DL) algorithms is challenging due to resource constraints on mobile and wearable devices, they provide performance improvements compared lightweight or shallow architectures. The widespread application areas for on-device DL include computer vision, image processing, natural language audio classification. However, sensing applications are also gaining attention. They can benefit from DL, given that these devices integrated with various sensors produce large amounts of data. This article reviews state-of-the-art studies particularly the sensor data analytics perspective. We first discuss general optimization techniques meet limitations devices. Then, we elaborate model update personalization review by classifying them according several aspects, including areas, sensors, types utilized algorithms, mode implementation, methods optimizing target training method, implementation toolkit/platform, metrics, consumption analysis. Finally, open issues future research directions about applications.

Language: Английский

Citations

33
Hybrid Solution Through Systematic Electrical Impedance Tomography Data Reduction and CNN Compression for Efficient Hand Gesture Recognition on Resource-Constrained IoT Devices DOI Creative Commons

Salwa Sahnoun,

Mahdi Mnif,

Bilel Ghoul

et al.

Future Internet, Journal Year: 2025, Volume and Issue: 17(2), P. 89 - 89

Published: Feb. 14, 2025

The rapid advancement of edge computing and Tiny Machine Learning (TinyML) has created new opportunities for deploying intelligence in resource-constrained environments. With the growing demand intelligent Internet Things (IoT) devices that can efficiently process complex data real-time, there is an urgent need innovative optimisation techniques overcome limitations IoT enable accurate efficient computations. This study investigates a novel approach to optimising Convolutional Neural Network (CNN) models Hand Gesture Recognition (HGR) based on Electrical Impedance Tomography (EIT), which requires signal processing, energy efficiency, real-time by simultaneously reducing input complexity using advanced model compression techniques. By systematically halving 1D CNN from 40 20 Boundary Voltages (BVs) applying method, we achieved remarkable size reductions 91.75% 97.49% BVs EIT inputs, respectively. Additionally, Floating-Point operations (FLOPs) are significantly reduced, more than 99% both cases. These have been with minimal loss accuracy, maintaining performance 97.22% 94.44% most significant result compressed model. In fact, at only 8.73 kB our demonstrates potential design strategies creating ultra-lightweight, high-performance CNN-based solutions near-full capabilities specifically case HGR inputs.

Language: Английский

Citations

1
Gesture Recognition Using MLP-Mixer With CNN and Stacking Ensemble for sEMG Signals DOI
Shu Shen,

Minglei Li,

Fan Mao

et al.

IEEE Sensors Journal, Journal Year: 2024, Volume and Issue: 24(4), P. 4960 - 4968

Published: Jan. 4, 2024

In recent years, gesture perception has become crucial to human–computer interaction (HCI) technologies. Among various techniques, recognition based on surface electromyography (sEMG) signals gained significant prominence, with deep-learning methods playing a pivotal role in this domain. However, as the demand for accurate continues rise, there is growing inclination toward selecting complex deep neural network architectures. This trend, however, poses challenges terms of performance and runtime requirements computing devices. article introduces novel method utilizing multilayer perceptron (MLP)-Mixer framework combined Stacking ensemble learning address these challenges. The proposed effectively captures features sEMG data by employing simple MLPs, achieving level accuracy comparable networks while simultaneously reducing inference time. Experimental results demonstrate that performs classification 80.03% 81.13% 49 actions open-source dataset NinaPro DB2, using window lengths 200 300 ms, respectively. Furthermore, achieves single speed 54.77 ms length ms. DB5, 250 presented rates 73.39% 74.82%, respectively, completing just 11.45 300-ms length. Notably, technique also demonstrates its ability mitigate impact individual differences accuracy.

Language: Английский

Citations

6
MCMP-Net: MLP combining max pooling network for sEMG gesture recognition DOI
Mian Xiang, Zhou Bingtao,

Cheng Shiqiang

et al.

Biomedical Signal Processing and Control, Journal Year: 2023, Volume and Issue: 90, P. 105846 - 105846

Published: Dec. 15, 2023

Language: Английский

Citations

11
Electromyography Signals in Embedded Systems: A Review of Processing and Classification Techniques DOI Creative Commons

José Félix Castruita-López,

Marcos Avilés, Diana C. Toledo-Pérez

et al.

Biomimetics, Journal Year: 2025, Volume and Issue: 10(3), P. 166 - 166

Published: March 10, 2025

This article provides an overview of the implementation electromyography (EMG) signal classification algorithms in various embedded system architectures. They address specifications used for different devices, such as number movements and type method. Architectures analyzed include microcontrollers, DSP, FPGA, SoC, neuromorphic computers/chips terms precision, processing time, energy consumption, cost. analysis highlights capabilities each technology real-time wearable applications smart prosthetics gesture control well importance local inference artificial intelligence models to minimize execution times resource consumption. The results show that choice device depends on required specifications, robustness model, be classified, limits knowledge concerning design budget. work a reference selecting technologies developing biomedical solutions based EMG.

Language: Английский

Citations

0
A Fast and Low-Impact Embedded Orientation Correction Algorithm for Hand Gesture Recognition Armbands DOI Creative Commons
Andrea Mongardi, Fábio Rossi, Andrea Prestia

et al.

Sensors, Journal Year: 2025, Volume and Issue: 25(7), P. 2188 - 2188

Published: March 30, 2025

Hand gesture recognition is a prominent topic in the recent literature, with surface ElectroMyoGraphy (sEMG) recognized as key method for wearable Human-Machine Interfaces (HMIs). However, sensor placement still significantly impacts systems performance. This study addresses displacement by introducing fast and low-impact orientation correction algorithm sEMG-based HMI armbands. The includes calibration phase to estimate armband real-time data correction, requiring only two distinct hand gestures terms of sEMG activation. ensures hardware database independence eliminates need model retraining, occurs prior classification or prediction. was implemented system featuring custom seven-channel an Artificial Neural Network (ANN) capable recognizing nine gestures. Validation demonstrated its effectiveness, achieving 93.36% average prediction accuracy arbitrary wearing orientation. also has minimal impact on power consumption latency, just additional 500 μW latency increase 408 μs. These results highlight algorithm's efficacy, general applicability, efficiency, presenting it promising solution electrode-shift issue applications.

Language: Английский

Citations

0
Low-Power High-Speed Neuromorphic Integrated Circuits DOI
Ramesh Kumar, Bal Chand Nagar

IGI Global eBooks, Journal Year: 2025, Volume and Issue: unknown, P. 235 - 262

Published: Feb. 28, 2025

The term neuromorphic is the biological activity of neurons. Neuromorphic integrated circuits (ICs) mimic characteristics nervous system. neural networks inspire IC creation in brain to construct hardware suitable for current need high-speed data processing. Memristor a fourth fundamental element proposed by Leon O. Chua 1971 and physically fabricated HP lab 2008. It most candidate computing. uses considerable promise synaptic systems providing energy-efficient memory processing capabilities. Understanding electrical properties neuronal membranes allows neurons retain their resting potential generate spikes. A neuron circuit provides underlying mechanisms action potentials, including signal production membrane function. This chapter describes implementation models applications spiking (SNNs). also highlights challenges designing ICs.

Language: Английский

Citations

0
Spiking neural networks for biomedical signal analysis DOI
Sang Ho Choi

Biomedical Engineering Letters, Journal Year: 2024, Volume and Issue: 14(5), P. 955 - 966

Published: July 5, 2024

Language: Английский

Citations

3