Scopus

Enhancing the Efficiency of Lightweight Deep Neural Networks in EEG-Based Emotion Recognition Using Early-Exit Techniques

Năm XB 2026 Tạp chí / Hội thảo Advances in Information and Communication Technology: Proceedings of the 4th International Conference, ICTA 2025, Volume 1 Volume 1, pp. 385-396 Đơn vị ICTU DOI / Link https://link.springer.com/chapter/10.1007/978-3-032-18316-3_43 ↗

Tác giả

Phung Trung Nghia

Tóm tắt

Emotion recognition from electroencephalogram signals (EEG-ER) plays an essential role in the development of intelligent human-machine interface applications. This paper proposes an enhancement to the existing EEG_SICNET architecture by integrating the Early-Exit (EE) mechanism, aiming to optimize performance on mobile and resource-constrained devices. The proposed model achieves higher classification accuracy (96.35% compared to 95.24%), while simultaneously reducing the number of parameters by a factor of 1.8 and improving inference speed by approximately 2.5 times on the DEAP dataset. Experimental results on additional benchmark datasets, namely DREAMER and AMIGOS, further validate the consistency of the model's effectiveness. Compared to recent state-of-the-art deep learning models designed for EEG-ER tasks, the proposed method demonstrates superior performance in terms of both accuracy and inference efficiency, reinforcing its potential for real-time deployment in practical scenarios.

Tài liệu tham khảo

[1] Nazar, M., Alam, M., Yafi, E., Mazliham, M.: A systematic review of human–computer interaction and explainable artificial intelligence in healthcare with artificial intelligence techniques. IEEE Access 1 (2021). https://doi.org/10.1109/ACCESS.2021.3127881

[2] Alarcao, S.M., Fonseca, M.J.: Emotions recognition using eeg signals: a survey. IEEE Trans. Affect. Comput. 10(3), 374–393 (2019). https://doi.org/10.1109/TAFFC.2017.2714671

[3] Sun, K., Wang, X., Zhao, Q.: A review of AIoT-based edge devices and lightweight deployment (2023)

[4] Thuong, D.T.M., Huy, N.P., Phung, T.-N., Cuong, D.N.: A lightweight deep learning network for emotion recognition applications on portable devices. Int. J. Knowl. Syst. Sci. 16(1), 1–23 (2025). https://doi.org/10.4018/IJKSS.373712

[5] Teerapittayanon, S., McDanel, B., Kung, H.T.: BranchyNet: fast inference via early exiting from deep neural networks. In: 2016 23rd International Conference on Pattern Recognition (ICPR), pp. 2464–2469. IEEE (2016). https://doi.org/10.1109/ICPR.2016.7900006

[6] Houssein, E.H., Hammad, A., Ali, A.A.: Human emotion recognition from EEG-based brain–computer interface using machine learning: a comprehensive review. Neural Comput. Appl. 34(15), 12527–12557 (2022). https://doi.org/10.1007/s00521-022-07292-4

[7] Guo, W., Xu, G., Wang, Y.: Multi-source domain adaptation with spatio-temporal feature extractor for EEG emotion recognition. Biomed. Signal Process. Control 84, 104998 (2023). https://doi.org/10.1016/j.bspc.2023.104998

[8] Li, J., et al.: Cross-subject EEG emotion recognition combined with connectivity features and meta-transfer learning. Comput. Biol. Med. 145, 105519 (2022). https://doi.org/10.1016/j.compbiomed.2022.105519

[9] Addad, Y., Lechervy, A., Jurie, F.: Seednet: simple early exiting dynamic neural networks without bells and whistles (2025). https://doi.org/10.2139/ssrn.5114735

[10] Matsubara, Y., Levorato, M., Restuccia, F.: Split computing and early exiting for deep learning applications: survey and research challenges. ACM Comput. Surv. 55(5), 1–30 (2023). https://doi.org/10.1145/3527155

[11] Lattanzi, E., Contoli, C., Freschi, V.: Do we need early exit networks in human activity recognition? Eng. Appl. Artif. Intell. 121, 106035 (2023). https://doi.org/10.1016/j.engappai.2023.106035

[12] Li, Y., Geller, T., Kim, Y., Panda, P.: SEENN: towards temporal spiking early-exit neural networks (2023). https://doi.org/10.48550/arXiv.2304.01230

[13] Wang, P., Guo, C., Xie, S., Qiao, X., Mao, L., Fu, X.: EEG emotion recognition based on knowledge distillation optimized residual networks. In: 2022 IEEE 6th Advanced Information Technology, Electronic and Automation Control Conference (IAEAC), pp. 574–581. IEEE (2022). https://doi.org/10.1109/IAEAC54830.2022.9929699

[14] Chao, H., Feng, S.: EEG emotion recognition based on self-distillation convolutional graph attention network. EAI Endors. Trans. e-Learn. 10 (2024). https://doi.org/10.4108/eetel.4974

[15] Akter, S., Prodhan, R.A., Pias, T.S., Eisenberg, D., Fresneda Fernandez, J.: M1M2: deep-learning-based real-time emotion recognition from neural activity. Sensors 22(21), 8467 (2022). https://doi.org/10.3390/s22218467

[16] Maheshwari, D., Ghosh, S.K., Tripathy, R.K., Sharma, M., Acharya, U.R.: Automated accurate emotion recognition system using rhythm-specific deep convolutional neural network technique with multi-channel EEG signals. Comput. Biol. Med. 134, 104428 (2021). https://doi.org/10.1016/j.compbiomed.2021.104428

[17] Lin, K., Zhang, L., Cai, J., Sun, J., Cui, W., Liu, G.: DSE-Mixer: a pure multilayer perceptron network for emotion recognition from EEG feature maps. J. Neurosci. Methods 401, 110008 (2024). https://doi.org/10.1016/j.jneumeth.2023.110008

Ghi chú

Source ID: 364; Author classification: ICTU - Main Author; First listed author: Duong Thi Mai Thuong, ; Contact author: Huy Nguyen Phuong; Total authors: 4; Springer matched from provided ICTA 2025 volumes

← Quay lại danh sách bài báo