Reading and Converting thoughts into Audible ...

Reading and Converting thoughts into Audible form based on BCI and EEG

Author Name : Masoud Alajmi, Shumukh Althgfi, Rawan Alshalwi, Tahani Alnasiri, Kholoud Almalki, Teaf Almalki, Lena Alzahrami, Nada Alotaibi, Atheer Alhashibri, Maha Aljuaid, Shada Alhumayani, Ftoun Alharthi

ABSTRACT Brain-Computer Interfaces (BCIs) enable direct engagement between neural mechanisms and external systems, offering potentially revolutionary solutions to support individuals with various disabilities. One of the main challenges for individuals with special needs is effective communication, particularly for those unable to use conventional methods. Electroencephalography (EEG)-based Brain-Computer Interfaces face obstacles such as low signal-to-noise ratio (SNR) and interference from environmental and physiological sources. Additionally, personalized models are needed due to the variation in brain signal patterns among individuals. Addressing these issues is essential for creating a reliable, efficient communication system for these individuals. This research aims to develop a novel system that facilitates communication for individuals with special needs by reading signals from the brain using an EEG device and analyzing them with the help of artificial intelligence (AI). EEG signals are captured and analyzed using deep learning models. Out of the 28 models evaluated on the dataset known as the “EEG dataset from the CVPR 2017 conference”, Fully Connected Neural Networks (FCNNs) demonstrated the best performance. This superiority is primarily attributed to their efficiency in extracting important features from the data, as well as their effective use of data augmentation techniques that help improve the model’s generalization ability. The proposed system, implemented on a Raspberry Pi, demonstrates the potential of AI and embedded systems in creating efficient communication tools that enhance independence and quality of life for users with communication challenges. Preliminary results demonstrate the effectiveness of this approach, indicating broad potential in medical, technical, and security applications, and establishing a new phase of interaction between humans and technology, in which the mind is the primary tool for control and expression.