Silent Speech Recognition by Surface Electromyography

Dobrucki, Andrzej; Pruchnicki, Piotr; Plaskota, Przemysław; Staroniewicz, Piotr; Brachmanski, Stefan; Walczyński, Maciej

doi:10.5772/60467

Cited by 5 publications

(3 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Audio signals collected in production halls -which are places with non-laboratory acoustic conditions -in addition to the component carrying useful information very often contain noise [3,4]. That noise component can be reduced using adaptive filtering algorithms [5,6].…”

Section: The First Section In Your Papermentioning

confidence: 99%

Digital system for sound sources location problem in multidimensional space using machine learning methods

Walczyński

Ruszkowski

Bożejko

2022

J. Phys.: Conf. Ser.

View full text Add to dashboard Cite

The issues presented in this work relate to the possibility of detecting the location of the sound source and its effectiveness in production conditions using recordings made with artificial head and processed with using an artificial neural networks. The use of an artificial head and artificial neural networks was motivated by an attempt to map human perception using available computer technology. The work attempts to map the features used for location human sources of sound through digital signal processing and machine learning. Artificial neural networks are commonly used, among others, in image recognition, where, using a camera or camera, the algorithm classifies objects in the image. Machine learning algorithms in addition allow for the implementation of self-learning speech or text recognition models. As part of the work, it was decided to localize the sound for a collection of fifteen different items sources relative to the head. To avoid duplicating similar measurements, the sound source was to the left of the artificial head. For signal processing and programming language was used to prepare the artificial neural network model Python with the Numpy numerical library, a library designed for signal processing Scipy, TensorFlow and Keras.

show abstract

Section: The First Section In Your Papermentioning

confidence: 99%

Digital system for sound sources location problem in multidimensional space using machine learning methods

Walczyński

Ruszkowski

Bożejko

2022

J. Phys.: Conf. Ser.

View full text Add to dashboard Cite

show abstract

“…Statistical model Hidden Markov Model (HMM) is one of the most classical recognition methods for both automatic speech recognition and silent speech recognition, and has been widely implemented in relevant studies (Meltzner et al, 2011(Meltzner et al, , 2017(Meltzner et al, , 2018Kubo et al, 2013). With the development of high-performance computers and data acquisition techniques, machine learning algorithms such as the basic Feedforward Neural Networks (Jong and Phukpattaranont, 2019), Linear Discriminant Analysis (LDA) (Liu et al, 2020), Bayes network (Dobrucki et al, 2016), Random Forests (RF) (Rameau, 2020;Zhang et al, 2020), and Support Vector Machine (SVM) (Rameau, 2020) have been used for silent speech recognition. Recently, deep learning has achieved great success in pattern recognition tasks, among which Convolutional Neural Network (CNN) shows outstanding performance not only in image classification but also in speech recognition (Liu et al, 2018;Xiong et al, 2018;Rashno et al, 2019).…”

Section: Introductionmentioning

confidence: 99%

A novel silent speech recognition approach based on parallel inception convolutional neural network and Mel frequency spectral coefficient

Zhang

Xie

et al. 2022

Front. Neurorobot.

View full text Add to dashboard Cite

Silent speech recognition breaks the limitations of automatic speech recognition when acoustic signals cannot be produced or captured clearly, but still has a long way to go before being ready for any real-life applications. To address this issue, we propose a novel silent speech recognition framework based on surface electromyography (sEMG) signals. In our approach, a new deep learning architecture Parallel Inception Convolutional Neural Network (PICNN) is proposed and implemented in our silent speech recognition system, with six inception modules processing six channels of sEMG data, separately and simultaneously. Meanwhile, Mel Frequency Spectral Coefficients (MFSCs) are employed to extract speech-related sEMG features for the first time. We further design and generate a 100-class dataset containing daily life assistance demands for the elderly and disabled individuals. The experimental results obtained from 28 subjects confirm that our silent speech recognition method outperforms state-of-the-art machine learning algorithms and deep learning architectures, achieving the best recognition accuracy of 90.76%. With sEMG data collected from four new subjects, efficient steps of subject-based transfer learning are conducted to further improve the cross-subject recognition ability of the proposed model. Promising results prove that our sEMG-based silent speech recognition system could have high recognition accuracy and steady performance in practical applications.

show abstract

“…The majority of speech recognition methods used with dysarthric patients is based on the analysis of acoustic signals [6], [7]. However, noisy environments are not suitable for this method because the acoustic signal is easily contaminated with ambient noise [8]. Hence, surface electromyography (sEMG) based speech recognition has been extensively proposed, especially in silent and non-audible environments [9].…”

Section: Introductionmentioning

confidence: 99%

Multimodal Data Fusion of Electromyography and Acoustic Signals for Thai Syllable Recognition

Jong

Herrera

Phukpattaranont

2021

IEEE J. Biomed. Health Inform.

View full text Add to dashboard Cite

Speech disorders such as dysarthria are common and frequent after suffering a stroke. Speech rehabilitation performed by a speech-language pathologist is needed to improve and recover. However, in Thailand, there is a shortage of speech-language pathologists. In this paper, we present a syllable recognition system, which can be deployable in a speech rehabilitation system to provide support to the limited speech-language pathologists available. The proposed system is based on a multimodal fusion of acoustic signal and surface electromyography (sEMG) collected from facial muscles. Multimodal data fusion is studied to improve signal collection under noisy situations while reducing the number of electrodes needed. The signals are simultaneously collected while articulating 12 Thai syllables designed for rehabilitation exercises. Several features are extracted from sEMG signals and five channels are studied. The best combination of features and channels is chosen to be fused with the mel-frequency cepstral coefficients extracted from the acoustic signal. The feature vector from each signal source is projected by spectral regression extreme learning machine and concatenated. Data from seven healthy subjects were collected for evaluation purposes. Results show that the multimodal fusion outperforms the use of a single signal source achieving up to ∼ 98% of accuracy. In other words, an accuracy improvement up to 5% can be achieved when using the proposed multimodal fusion. Moreover, its low standard deviations in classification accuracy compared to those from the unimodal fusion indicate the improvement in the robustness of the syllable recognition.

show abstract

Silent Speech Recognition by Surface Electromyography

Cited by 5 publications

References 17 publications

Digital system for sound sources location problem in multidimensional space using machine learning methods

Digital system for sound sources location problem in multidimensional space using machine learning methods

A novel silent speech recognition approach based on parallel inception convolutional neural network and Mel frequency spectral coefficient

Multimodal Data Fusion of Electromyography and Acoustic Signals for Thai Syllable Recognition

Contact Info

Product

Resources

About