Intelligent Autonomous Driving Assistance system based on Brain Computer Interface

Advancements in neurotechnology and signal processing have revolutionized Brain-Computer Interfaces (BCIs), enabling direct and functional connectivity between the human brain and computational systems. Particularly, Steady-State Visual Evoked Potentials (SSVEP) and reinforcement learning algorithms have garnered significant attention for their potential in creating more efficient and user-friendly BCIs. Utilizing SSVEP in Brain-Computer Interfaces negates the need for extensive subject training, a notable advancement particularly beneficial in applications such as assistive communication and rehabilitation.

This approach ensures high accuracy and information transfer rates, making it highly effective in varied social and therapeutic settings where user-friendly and adaptable interfaces are crucial.

This study develops a SSVEP brain-computer interface, integrated with a Unity-based virtual environment, to enable control of a virtual car. The focus is on harnessing SSVEP-BCI for intuitive interaction within a simulated Unity environment. We use the common spatial pattern algorithm to calculate the best feature vector for the EEG data and then train it in Linear Discriminant Analysis, Muti-layer perceptron classifier and Support Vector Machine (SVM) classifiers, with the SVM giving the best classification accuracy at 89.7%. This paper serves to introduce ongoing work and future applications of SSVEPBCI in controlling real-life assistive devices to help disabled people regain their physical mobility and provide opportunities for independent living.