Cueless EEG imagined speech achieves 97.93% accuracy for subject identification

Researchers have developed a novel method for subject identification using electroencephalogram (EEG) signals from imagined speech, achieving a remarkable 97.93% accuracy. This approach eliminates the need for external visual or auditory cues, requiring subjects to imagine words from a predefined list. The study utilized a dataset of over 4,350 trials from 11 subjects and evaluated various classification techniques, including traditional machine learning models like SVM and XGBoost, alongside advanced deep learning architectures such as EEG Conformer and Shallow ConvNet. The findings suggest significant potential for cueless EEG-based imagined speech in applications like brain-computer interfaces. AI

IMPACT This research could lead to more secure and private biometric identification systems, particularly for brain-computer interfaces.

RANK_REASON The item is an academic paper detailing a new dataset and benchmarks for a specific AI research area (EEG-based imagined speech for subject identification). [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.AI →

• Ali Derakhshesh
• alphaXiv
• arXiv
• brain-computer interfaces
• CatalyzeX
• DagsHub
• electroencephalogram
• Gotit.pub
• Hugging Face
• ScienceCast
• Shallow ConvNet
• support vector machine
• XGBoost

• paper
• other

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →