Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Toward Practical BCI: A Real-time Wireless Imagined Speech EEG Decoding System

Ji-Ha Park, Heon-Gyu Kwak, Gi-Hwan Shin, Yoo-In Jeon, Sun-Min Park, Ji-Yeon Hwang, Seong-Whan Lee

TL;DR

This work tackles the gap between laboratory BCI demonstrations and practical daily use by presenting a real-time, wireless imagined-speech EEG decoding system that incorporates user identification for personalized models and real-time data streaming via Lab Streaming Layer. The authors employ a dual sequential transformer encoder to decode imagined-speech commands from both wired (32-channel) and wireless (12-channel) EEG setups, with a lightweight calibration pipeline to tailor the decoder to individual users. In experiments with three participants across four imagined commands, the wired device achieved 62% accuracy (macro-F1 0.62), while a portable wireless headset reached 46.67% accuracy (macro-F1 0.46), both above chance. The results demonstrate the feasibility of practical, real-time BCI across hardware platforms and outline a path toward robust, user-specific neural interfaces for real-world use.

Abstract

Brain-computer interface (BCI) research, while promising, has largely been confined to static and fixed environments, limiting real-world applicability. To move towards practical BCI, we introduce a real-time wireless imagined speech electroencephalogram (EEG) decoding system designed for flexibility and everyday use. Our framework focuses on practicality, demonstrating extensibility beyond wired EEG devices to portable, wireless hardware. A user identification module recognizes the operator and provides a personalized, user-specific service. To achieve seamless, real-time operation, we utilize the lab streaming layer to manage the continuous streaming of live EEG signals to the personalized decoder. This end-to-end pipeline enables a functional real-time application capable of classifying user commands from imagined speech EEG signals, achieving an overall 4-class accuracy of 62.00 % on a wired device and 46.67 % on a portable wireless headset. This paper demonstrates a significant step towards truly practical and accessible BCI technology, establishing a clear direction for future research in robust, practical, and personalized neural interfaces.

Toward Practical BCI: A Real-time Wireless Imagined Speech EEG Decoding System

TL;DR

This work tackles the gap between laboratory BCI demonstrations and practical daily use by presenting a real-time, wireless imagined-speech EEG decoding system that incorporates user identification for personalized models and real-time data streaming via Lab Streaming Layer. The authors employ a dual sequential transformer encoder to decode imagined-speech commands from both wired (32-channel) and wireless (12-channel) EEG setups, with a lightweight calibration pipeline to tailor the decoder to individual users. In experiments with three participants across four imagined commands, the wired device achieved 62% accuracy (macro-F1 0.62), while a portable wireless headset reached 46.67% accuracy (macro-F1 0.46), both above chance. The results demonstrate the feasibility of practical, real-time BCI across hardware platforms and outline a path toward robust, user-specific neural interfaces for real-world use.

Abstract

Brain-computer interface (BCI) research, while promising, has largely been confined to static and fixed environments, limiting real-world applicability. To move towards practical BCI, we introduce a real-time wireless imagined speech electroencephalogram (EEG) decoding system designed for flexibility and everyday use. Our framework focuses on practicality, demonstrating extensibility beyond wired EEG devices to portable, wireless hardware. A user identification module recognizes the operator and provides a personalized, user-specific service. To achieve seamless, real-time operation, we utilize the lab streaming layer to manage the continuous streaming of live EEG signals to the personalized decoder. This end-to-end pipeline enables a functional real-time application capable of classifying user commands from imagined speech EEG signals, achieving an overall 4-class accuracy of 62.00 % on a wired device and 46.67 % on a portable wireless headset. This paper demonstrates a significant step towards truly practical and accessible BCI technology, establishing a clear direction for future research in robust, practical, and personalized neural interfaces.

Paper Structure

This paper contains 14 sections, 2 figures, 1 table.

Table of Contents

  1. INTRODUCTION
  2. MATERIALS AND METHODS
  3. Data Acquisition
  4. Experimental setup
  5. EEG recording and preprocessing
  6. Decoding Model
  7. Personalized BCI Application
  8. User identification
  9. Intention classification
  10. System implementation
  11. RESULTS AND DISCUSSION
  12. Performance on Wired EEG Device
  13. Performance on Wireless EEG Device
  14. CONCLUSIONS

Figures (2)

  • Figure 1: Overall framework of a real-time practical BCI system. Imagined speech EEG signals are acquired from either a wired (Brain Products) or wireless (Emotiv EPOC X) device. The user identifier module first processes the signals to recognize the user's information. Based on this identity, a corresponding pre-trained model is loaded and subsequently fine-tuned using a small set of newly acquired calibration data. This fine-tuned model then serves as the intention decoder. For real-time operation, live EEG signals are streamed via LSL directly to the personalized intention decoder, which performs inference to classify the user's intention for the final application display.
  • Figure 2: Overview of the proposed BCI application workflow. Users can select either a wired or wireless device for EEG signal input. The main GUI is used to select functions such as user registration, calibration, or system activation. The system is designed to visualize EEG signals in real-time and display the corresponding inference results.