International Journal of Innovative Research in Computer and Communication Engineering
ISSN Approved Journal | Impact factor: 8.771 | ESTD: 2013 | Follows UGC CARE Journal Norms and Guidelines
| Monthly, Peer-Reviewed, Refereed, Scholarly, Multidisciplinary and Open Access Journal | High Impact Factor 8.771 (Calculated by Google Scholar and Semantic Scholar | AI-Powered Research Tool | Indexing in all Major Database & Metadata, Citation Generator | Digital Object Identifier (DOI) |
| TITLE | Deep Learning Based EEG Signal Classification for Brain-Computer Interface Applications Using Neurobalancenet |
|---|---|
| ABSTRACT | Motor-imagery (MI) based Brain-Computer Interface (BCI) systems built on Electroencephalography (EEG) are gaining traction in neuro-rehabilitation, assistive computing and human-machine interaction, yet their real-world accuracy is routinely limited by class imbalance, inter-trial variability and scarce labelled recordings. Lightweight architectures such as EEGNet are computationally attractive but tend to under-recognise minority movement classes, which drags down overall decoding performance. This paper proposes NeuroBalanceNet, an EEG classification pipeline that pairs a compact convolutional backbone with an Efficient Channel Attention (ECA) module and a Focal-Loss training objective, so that difficult and under-represented trials receive proportionally greater weight during learning, together with Gaussian-noise and MixUp-based augmentation to enlarge the effective training set. Because large-scale, four-class motor-imagery corpora such as BCI Competition IV-2a are costly to acquire, the pipeline is first validated end-to-end on a controlled synthetic multichannel EEG dataset spanning three activation levels, using a 1-D convolutional network with a temporal-attention head as the working prototype of the attention-guided branch of NeuroBalanceNet. The prototype is assessed using accuracy, class-wise precision/recall/F1 and training-validation convergence curves; the observed behaviour - strong recognition of the extreme classes, weaker recognition of the overlapping middle class, and a marked synthetic-to-real accuracy gap - is used to justify the design choices that NeuroBalanceNet is expected to carry over to genuine motor-imagery decoding. The paper concludes with a roadmap for porting the validated pipeline onto the BCI Competition IV-2a dataset. |
| AUTHOR | G. SWATHI, CH. HARIKA Department of Computer Science and Engineering, St. Mary's Women's Engineering College, Budampadu, Guntur, Andhra Pradesh, India |
| VOLUME | 186 |
| DOI | DOI: 10.15680/IJIRCCE.2026.1407024 |
| pdf/24_Deep Learning Based EEG Signal Classification for Brain-Computer Interface Applications Using Neurobalancenet.pdf | |
| KEYWORDS | |
| References | [1] A. L. Goldberger et al., "PhysioBank, PhysioToolkit, and PhysioNet: Components of a new research resource for complex physiologic signals," Circulation, 2000. [2] Y. Roy, H. Banville, I. Albuquerque, A. Gramfort, T. H. Falk, and J. Faubert, "Deep learning-based electroencephalography analysis: a systematic review," Journal of Neural Engineering, 2020. [3] Z. Zhang, C. C. K. Cheung, and S. Kwong, "Data augmentation for EEG-based emotion recognition using generative adversarial networks," IEEE Trans. Affective Computing, 2021. [4] S. Makeig et al., "Dynamic brain sources of visual evoked responses," Science, 2002. [5] F. Lotte et al., "A review of classification algorithms for EEG-based brain-computer interfaces: a 10 year update," Journal of Neural Engineering, 2018. [6] R. T. Schirrmeister et al., "Deep learning with convolutional neural networks for EEG decoding and visualization," Human Brain Mapping, 2017. [7] J. Roy et al., "Challenges and opportunities in EEG signal classification: A review," Biomedical Signal Processing and Control, 2019. [8] P. Bashivan, I. Rish, M. Yeasin, and N. Codella, "Learning representations from EEG with deep recurrent-convolutional neural networks," ICLR, 2016. [9] S. Roy, M. Kiral-Kornek, and S. Harrer, "ChronoNet: A deep recurrent neural network for abnormal EEG identification," ICASSP, 2019. [10] Y. Banville et al., "Uncovering the structure of clinical EEG signals with deep convolutional neural networks," Scientific Reports, 2021. [11] D. Roy et al., "Data augmentation for EEG-based mental workload classification," IEEE Trans. Neural Systems and Rehabilitation Engineering, 2022. [12] A. Vaswani et al., "Attention is all you need," Advances in Neural Information Processing Systems, vol. 30, pp. 5998-6008, 2017. [13] V. J. Lawhern, A. J. Solon, N. R. Waytowich, S. M. Gordon, C. P. Hung, and B. J. Lance, "EEGNet: A compact convolutional neural network for EEG-based brain-computer interfaces," Journal of Neural Engineering, vol. 15, no. 5, 2018. [14] X. Zhang, L. Yao, Y. Zhang, X. Wang, and T. Shen, "MST-AttnNet: A multi-scale temporal attention network for EEG-based emotion recognition," IEEE Trans. Affective Computing, 2021. |
