A Novel Signal-to-Image Conversion Approach with Ensembles of Pretrained CNNs for Motor Imagery EEG Signals DOI
Çağatay Murat Yılmaz,

Bahar Hatipoglu Yilmaz,

Cemal Köse

и другие.

Опубликована: Сен. 20, 2023

Deep learning techniques can recognize and learn significant features automatically. The advantages of deep should also be used to solve problems in other fields. Therefore, we concentrated on classifying motor imagery EEG signals with transfer learning. present study first proposed an signal-to-image conversion method that uses 2D signal representations. Thanks this, employed images as input instead standard one-dimensional features. Second, AlexNet, GoogLeNet, SqueezeNet pre-trained convolutional neural network frameworks for classification. We hard voting merged the outputs networks. performed experiments dataset 2a BCI Competition IV, achieving 84.18±5.37% classification accuracy 0.66±0.12 kappa across nine subjects, even limited training data. results showed has potential, particularly processing biomedical signals.

Язык: Английский

MS-HyFS: A novel multiscale hybrid framework with Scalable electrodes for motor imagery classification DOI

Ziheng Guo,

Yuan Feng,

Ming Ma

и другие.

Biomedical Signal Processing and Control, Год журнала: 2025, Номер 106, С. 107706 - 107706

Опубликована: Фев. 21, 2025

Язык: Английский

Процитировано

0

A Novel Signal-to-Image Conversion Approach with Ensembles of Pretrained CNNs for Motor Imagery EEG Signals DOI
Çağatay Murat Yılmaz,

Bahar Hatipoglu Yilmaz,

Cemal Köse

и другие.

Опубликована: Сен. 20, 2023

Deep learning techniques can recognize and learn significant features automatically. The advantages of deep should also be used to solve problems in other fields. Therefore, we concentrated on classifying motor imagery EEG signals with transfer learning. present study first proposed an signal-to-image conversion method that uses 2D signal representations. Thanks this, employed images as input instead standard one-dimensional features. Second, AlexNet, GoogLeNet, SqueezeNet pre-trained convolutional neural network frameworks for classification. We hard voting merged the outputs networks. performed experiments dataset 2a BCI Competition IV, achieving 84.18±5.37% classification accuracy 0.66±0.12 kappa across nine subjects, even limited training data. results showed has potential, particularly processing biomedical signals.

Язык: Английский

Процитировано

2