Intention Recognition Technology

The goal of this project is to set up a technology for controlling external application using human implicit intention.

ETC

Intelligence

v
January 18, 2025

Runtime Environment

  • Windows 10 64bit
  • SSVEP, ERP
    • 27-inch LCD monitor, 60Hz
    • Biosemi ActiveTwo
  • Drowsiness detection
    • Empatica E4

Development Environment

o    Eclipse Java EE IDE for Web Developers

o    JDK

o    Python3 installation

o    JAVA – Android Studio installation

Get Started

  • SSVEP / ERP
    • Open the Biosemi Labview and setting the option(sampling rate, channel, etc.)
    • Setting the parameter of python / Java code(trial, stimulus time, etc.)
    • Start the Boisemi Labview
    • Execute python / Java program
  • Drowsiness detection
    • Wrist wear and power on the Empatica E4 device
    • Connection with a PPG measurement program
    • Button control of the PPG measurement program
    • Offline: For storing PPG data
    • Online: Real-time PPG measurement and drowsiness recognition

 

Overall description of source code

  • SSVEP / ERP
    • Creating the visual stimulus of SSVEP / ERP
    • Acquisition of EEG data of SSVEP / ERP
    • Preprocessing the EEG data of SSVEP / ERP acquired.
    • Feature Extraction (Common Spatial Pattern, Canonical Correlation Analysis, etc.)
    • Classification (Linear Discriminant Analysis, Support Vector Machine)

  

  • Drowsiness detection
    • Acquisition of user’s PPG data using PPG measuring device (‘Empatica E4’)
    • Real-time communication of acquired PPG data
    • Classifier model training using acquired PPG data (offline)
    • User state classification using PPG data acquired in real time (online)

 

The development of Intention Recognition Technology based on Brain Signals


Introduction

  • The goal of this project is to set up a technology for controlling external application using human implicit intention.
  • Data from the EEG extraction device is transmitted by TCP/IP protocol.
  • We use the EEG-based BCI paradigms to control IoT device.
  • Among the BCI paradigms, we use Steady-State Visual Evoked Potential (SSVEP) and Event-related potential (ERP) to recognize human intentions.
  • We developed the on/off control technology of Philips Hue lamp through human drowsiness state recognition based on PPG.
  • The IoT device used in this technology are Philips Hue lamp, Belkin Wemo Plug, etc.