MemEye

Overview

Memeye is a real-time bio-signal processing system that accurately identifies human learning phases by combining pupil dilation data and electrodermal activity. The system processes multiple bio-signals from pupil labs and emotibit simultaneously to understand a learner’s cognitive state, enabling adaptive learning systems to automatically adjust content difficulty.

Role & Contributions

• Led the development of an end-to-end inference pipeline for real-time bio-signal processing
• Designed and implemented signal processing algorithms for noise reduction and feature extraction
• Created a multi-modal fusion architecture to combine insights from pupil data and electrodermal activity
• Built an efficient real-time processing system that runs on standard hardware

Key Findings

• Achieved 98% accuracy in learning phase classification through multi-modal bio-signal analysis
• Demonstrated that combining pupil dilation and electrodermal activity provides more reliable classification than single-source data
• Established that real-time processing of bio-signals is feasible for practical applications
• Validated the system’s ability to detect subtle changes in cognitive states during learning

Impact & Implications

• Currently preparing two papers from this research:
  • A dataset paper documenting the collection and analysis of multi-modal bio-signal data during learning
  • An experimental paper focusing on the system’s methodology and classification results
• Opens new possibilities for personalized learning experiences based on cognitive state
• Provides a foundation for developing more sophisticated human-AI interaction systems
• Demonstrates the potential of multi-modal bio-signal processing for understanding human cognitive states

To know more about this work: Code