Gesture recognition based on ConvLSTM-attention implementation of small data sEMG signals

Cao, Xin; Iwase, Masami; Inoue, Jun; Maeda, Eriko

doi:10.1145/3341162.3343835

Cited by 7 publications

(3 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Ding et al [19] used a parallel multiple-scale CNN for hand gesture classification, while Wei et al [20] employed a CNN with multiple sub-streams. Gao et al [21] proposed a dualflow network with CNN and LSTM, Wu et al [22] used CNN and LSTM with attention mechanism, Xie et al [23] combined CNN and LSTM, and Tong et al [24] used CNN and RNN for gesture classification. Tsinganos et al [25] achieved improved performance with a temporal convolutional network (TCN) on Ninapro DB1, and Zanghieri et al [26] developed TEMPONet, a TCN-based network on an embedded system, outperforming existing methods on Ninapro DB6.…”

Section: E Mixed Network Structuresmentioning

confidence: 99%

LSTM-MSA: A Novel Deep Learning Model With Dual-Stage Attention Mechanisms Forearm EMG-Based Hand Gesture Recognition

Zhang,

Qu,

Teng

et al. 2023

IEEE Trans. Neural Syst. Rehabil. Eng.

View full text Add to dashboard Cite

This paper introduces the Long Short-Term Memory with Dual-Stage Attention (LSTM-MSA) model, an approach for analyzing electromyography (EMG) signals. EMG signals are crucial in applications like prosthetic control, rehabilitation, and human-computer interaction, but they come with inherent challenges such as non-stationarity and noise. The LSTM-MSA model addresses these challenges by combining LSTM layers with attention mechanisms to effectively capture relevant signal features and accurately predict intended actions. Notable features of this model include dual-stage attention, end-to-end feature extraction and classification integration, and personalized training. Extensive evaluations across diverse datasets consistently demonstrate the LSTM-MSA's superiority in terms of F1 score, accuracy, recall, and precision. This research provides a model for real-world EMG signal applications, offering improved accuracy, robustness, and adaptability.

show abstract

Section: E Mixed Network Structuresmentioning

confidence: 99%

LSTM-MSA: A Novel Deep Learning Model With Dual-Stage Attention Mechanisms Forearm EMG-Based Hand Gesture Recognition

Zhang,

Qu,

Teng

et al. 2023

IEEE Trans. Neural Syst. Rehabil. Eng.

View full text Add to dashboard Cite

show abstract

“…According to Cao et al (2019) [84], hybrid CNN and LSTM models can successfully be used for improving the performance of LSTM models on movement classification tasks. They show that the usage of a single technique usually is not enough and requires a huge amount of data for training purposes, which is usually not widely available.…”

Section: Machine Learning For Biological Data Processingmentioning

confidence: 99%

“…Future work on EMG signal prediction could incorporate convolution layers (Conv1D) before the MLP or before the LSTM networks, to capture new spatial patterns from the time-series data, such as those included in [84,89,115]. We can also propose using these prediction models as inputs for other types of NN that can optimize the control parameters of FES devices.…”

Section: Chapter 7 Conclusionmentioning

confidence: 99%

Parkinson EMG signal prediction and generation with neural networks

Piccolli¹

View full text Add to dashboard Cite

I want to thank UNICAMP and, in particular, to the entire team of the Institute of Computing and the Dean of Graduate Studies for the opportunity to develop my master's degree within the Institution, which provided the environment, resources, and professors for my personal development during my master's degree.I would like to especially thank my advisor, professor Esther L. Colombini, for her tireless dedication, contribution, and support since the beginning of my master's degree. She always tried to help and guide me at all stages, even from a remote location and in different time zones, looking for alternatives and new ideas even when I seemed to be lost, reviewing, and supporting my work even in the most challenging deadlines. I am eternally grateful for her support, without which this work and the published articles would not exist.I also thank professor Dra. Maria Claudia Ferrari de Castro, for collaborating with ideas and references, for participating in my Qualification round, and above all, for sharing the Parkinson's Disease datasets from her previous work, without which this work would also not be possible.I am thankful to my Siemens co-workers in Brazil and Germany, for the opportunity to share ideas, problems and solutions, and for allowing me to dedicate myself to the master's degree simultaneously with my career activities.Finally, I would like to thank my whole family for their support and encouragement since the beginning of this journey. To my parents, for believing in me and for always being proud, even in the smallest achievements, to my wife, who during all this time faced countless challenges, such as relocating to another country, having the birth of our first child, and despite all that, still finding the time to encourage me and providing the best environment so I could focus on this work. Finally, I'm grateful to my son Erik, who has brought me a new meaning to life and inspiration every day.

show abstract

Surface Electromyography Signal Recognition Based on Deep Learning for Human-Robot Interaction and Collaboration

Mendes

2022

J Intell Robot Syst

View full text Add to dashboard Cite

Gesture recognition based on ConvLSTM-attention implementation of small data sEMG signals

Cited by 7 publications

References 17 publications

LSTM-MSA: A Novel Deep Learning Model With Dual-Stage Attention Mechanisms Forearm EMG-Based Hand Gesture Recognition

LSTM-MSA: A Novel Deep Learning Model With Dual-Stage Attention Mechanisms Forearm EMG-Based Hand Gesture Recognition

Parkinson EMG signal prediction and generation with neural networks

Surface Electromyography Signal Recognition Based on Deep Learning for Human-Robot Interaction and Collaboration

Contact Info

Product

Resources

About