TransTCNet: Transformer-Based Temporal-Contextual Network for Low-Latency Typing Interfaces on Edge Devices
Source abstract
Abstract
A distinct typing interface using surface electromyography (sEMG) can facilitate silent, hands-free typing by interpreting muscle activity in relation to specific keystrokes. Character-level recognition poses greater challenges than coarse gesture recognition because it is sensitive to subtle temporal variations and overlapping muscle dynamics. Temporal features are essential for typing recognition because keypresses may differ in duration, force, and accompanying hand movements across users. This paper proposes TransTCNet, a two-stage deep neural network architecture with a causal convolutional layer for learning local features and a transformer-based component for learning long-range temporal interactions. We evaluated our network on a publicly available 26-class typing sEMG dataset acquired from 19 individuals. The model achieved a validation accuracy of 96.53%, exceeding the baseline models. Our study revealed generalization among participants, and the AUC values were also high (>0.994) across all classes. The model was highly reliable and exhibited high prediction confidence (>0.9), enabling us to achieve a high training accuracy (97.86%) for real-time filtering decisions. TransTCNet could be suitable for wearable and edge devices due to its efficient architecture and low inference cost. The model’s ability to consistently decode fine-grained neuromuscular signals across users makes it well-suited for real-time applications such as adaptive user interfaces, virtual and augmented reality, prosthetic control, and communication systems.
Source: crossref · Quality: baseline
Research context
OpenAlex topics: Muscle activation and electromyography studies · Advanced Sensor and Energy Harvesting Materials · Interactive and Immersive Displays
SciVX discovery tags are terminology-based. OpenAlex topics are source-provided classifications.
Affiliations
Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences, Shenzhen 518055, China
Institute of Ultrasonic Technology, Shenzhen Polytechnic University, Shenzhen 518055, China
Department of Computer Science and Engineering, School of Engineering, Nanfang College, Guangzhou 510970, China
Funding
Postdoctoral Foundation · Research Projects of Department of Education of Guangdong Province · Research Projects of Shenzhen Polytechnic University
References (37)
Show source-provided references
1. Yeo, A., Kwok, B.W., Joshna, A., Chen, K., and Lee, J.S. (2024). Entering the next dimension: A review of 3d user interfaces for virtual reality. Electronics, 13.10.3390/electronics13030600 ↗
2. HeadTrack: Real-Time Human–Computer Interaction via Wireless Earphones10.1109/JSAC.2023.3345381 ↗
3. Surface electromyography as a natural human–machine interface: A review10.1109/JSEN.2022.3165988 ↗
4. Zaim, T., Abdel-Hadi, S., Mahmoud, R., Khandakar, A., Rakhtala, S.M., and Chowdhury, M.E. (2025). Machine Learning-and Deep Learning-Based Myoelectric Control System for Upper Limb Rehabilitation Utilizing EEG and EMG Signals: A Systematic Review. Bioengineering, 12.10.3390/bioengineering12020144 ↗
5. Eddy, E., Campbell, E., Bateman, S., and Scheme, E. (2024). Big data in myoelectric control: Large multi-user models enable robust zero-shot EMG-based discrete gesture recognition. Front. Bioeng. Biotechnol., 12.10.3389/fbioe.2024.1463377 ↗
6. EMG dataset for gesture recognition with arm translation10.1038/s41597-024-04296-8 ↗
7. A wearable human–machine-interface (HMI) system based on colocated EMG-pFMG sensing for hand gesture recognition10.1109/TMECH.2024.3386929 ↗
8. Ma, C., Wang, C., Zhu, D., Chen, M., Zhang, M., and He, J. (2025). The Investigation of the Relationship Between Individual Pain Perception, Brain Electrical Activity, and Facial Expression Based on Combined EEG and Facial EMG Analysis. J. Pain Res., 21–32.10.2147/JPR.S477658 ↗
9. Comparative analysis of electrical signals in facial expression muscles
10. Ullah, A., Zhang, F., Song, Z., Wang, Y., Zhao, S., Riaz, W., and Li, G. (2024). Surface Electromyography-Based Recognition of Electronic Taste Sensations. Biosensors, 14.10.3390/bios14080396 ↗
11. Gender Differences in Taste Sensations Based on Frequency Analysis of Surface Electromyography10.1177/00315125231169882 ↗
12. Salkanovic, A., Sušanj, D., Batistić, L., and Ljubic, S. (2025). Beyond Signatures: Leveraging Sensor Fusion for Contextual Handwriting Recognition. Sensors, 25.10.20944/preprints202503.1018.v1 ↗
13. Tigrini, A., Ranaldi, S., Verdini, F., Mobarak, R., Scattolini, M., Conforto, S., Schmid, M., Burattini, L., Gambi, E., and Fioretti, S. (2024). Intelligent human–computer interaction: Combined wrist and forearm myoelectric signals for handwriting recognition. Bioengineering, 11.10.3390/bioengineering11050458 ↗
14. Electromyographic typing gesture classification dataset for neurotechnological human-machine interfaces10.1038/s41597-025-04763-w ↗
15. Enhanced complex human activity recognition system: A proficient deep learning framework exploiting physiological sensors and feature learning10.1109/LSENS.2023.3326126 ↗
16. Temporal-channel convolution with self-attention network for human activity recognition using wearable sensors10.1016/j.knosys.2023.110867 ↗
17. Singh, S., Choudhury, N.A., and Soni, B. (2023, January 24–25). Gait recognition using activities of daily livings and ensemble learning models. Proceedings of the International Conference on Advances in IoT and Security with AI, New Delhi, India.10.1007/978-981-99-5085-0_20 ↗
18. Zhang, S., Li, Y., Zhang, S., Shahabi, F., Xia, S., Deng, Y., and Alshurafa, N. (2022). Deep learning in human activity recognition with wearable sensors: A review on advances. Sensors, 22.10.3390/s22041476 ↗
19. Human activity recognition using inertial, physiological and environmental sensors: A comprehensive survey10.1109/ACCESS.2020.3037715 ↗
20. Choudhury, N.A., and Soni, B. (2022, January 1–3). An efficient CNN-LSTM approach for smartphone sensor-based human activity recognition system. Proceedings of the 2022 5th International Conference on Computational Intelligence and Networks (CINE), Odisha, India.10.1109/CINE56307.2022.10037495 ↗
21. Estimating finger joint angles by surface EMG signal using feature extraction and transformer-based deep learning model10.1016/j.bspc.2023.105447 ↗
22. Riaz, W., Ullah, A., and Ji, J. (2025). Multi-Scale Attention Networks with Feature Refinement for Medical Item Classification in Intelligent Healthcare Systems. Sensors, 25.10.3390/s25175305 ↗
23. TPRO-NET: An EEG-based emotion recognition method reflecting subtle changes in emotion10.1038/s41598-024-62990-4 ↗
24. Riaz, W., Ji, J., and Ullah, A. (2025). TriViT-Lite: A Compact Vision Transformer–MobileNet Model with Texture-Aware Attention for Real-Time Facial Emotion Recognition in Healthcare. Electronics, 14.10.3390/electronics14163256 ↗
25. Ullah, A., Song, Z., Riaz, W., Qi, X., and Hossain, M.M. (2026). Hand Gesture-Based Biometric Verification and Identification Using Embedded-STQNet Deep Neural Architecture in Security-Oriented Systems. IEEE Internet Things J.10.1109/JIOT.2026.3675359 ↗
26. Multi-ResAtt: Multilevel residual network with attention for human activity recognition using wearable sensors
27. Hnoohom, N., Jitpattanakul, A., You, I., and Mekruksavanich, S. (2021, January 1–3). Deep learning approach for complex activity recognition using heterogeneous sensors from wearable device. Proceedings of the 2021 Research, Invention, and Innovation Congress: Innovation Electricals and Electronics (RI2C), Bangkok, Thailand.10.1109/RI2C51727.2021.9559773 ↗
28. Physique-based human activity recognition using ensemble learning and smartphone sensors10.1109/JSEN.2021.3077563 ↗
29. An efficient and lightweight deep learning model for human activity recognition on raw sensor data in uncontrolled environment10.1109/JSEN.2023.3312478 ↗
30. TASTA: Text-Assisted Spatial and Temporal Attention Network for Video Question Answering10.1002/aisy.202200131 ↗
31. Pradhan, A. (2024). Electromyography-Based Biometrics for Secure and Robust Personal Identification and Authentication. [Ph.D. Thesis, University of Waterloo].
32. Ullah, A., Song, Z., Riaz, W., and Wang, Y. (2025). GTMH-TasteNet: Advanced Deep Learning for sEMG-Based Taste Sensation Recognition. Tsinghua Sci. Technol.10.26599/TST.2025.9010143 ↗
33. Simple recognition of hand gestures using single-channel EMG signals10.1177/09544119231225528 ↗
34. Liu, G., Wang, Z., Xi, C., He, Z., Guo, S., Zhang, R., and Yao, D. (2024). Kinetic and Kinematic Sensors-free Approach for Estimation of Continuous Force and Gesture in sEMG Prosthetic Hands. arXiv.
35. Breast cancer prediction by ensemble meta-feature space generator based on deep neural network10.1016/j.bspc.2023.105382 ↗
36. Wetzel, S.J., Ha, S., Iten, R., Klopotek, M., and Liu, Z. (2025). Interpretable machine learning in physics: A review. arXiv.
37. An LSTM Network With Neural Plasticity for Driver Fatigue Recognition on Real Roads10.1109/TIE.2025.3585046 ↗