SEPICS PROGRAM Implementation of an emotional detection module. 1.

1

SEPICS PROGRAM

Implementation of an emotional detection module.

2

CONTENT

Presentation of the student

Presentation of the HERON laboratory

Why choosing SEPCIS?

Presentation of internships project

Project progress

Development plan

3

PRESENTATION OF THE STUDENT

Postgraduate student from University of

Glamorgan

Bachelor’s degree Computer Science

Masters degree Intelligent Computer Systems

4

PRESENTATION OF THE HERON LABORATORY

HERON – Higher Educational Research ON

tutoring systems.

Established in 1986.

Includes researches from several universities.

Altogether around 40 members.

First team in Quebec to specialize in this research

area.

Recognized internationally.

Received several grants.

5

HERON FIELD OF EXPERTISE

Artificial intelligence

Emotional intelligence

Intelligent tutoring systems (ITS)

Intelligent agents

EEG-based education

6

WHY CHOOSE SEPICS?

SEPICS – Student Exchange Program in

Intelligent Computer Systems.

Opportunity to work on a research project

regarding intelligent systems.

A chance to further my academic knowledge.

Travelling to Canada.

Experience different lifestyle and culture.

7

INTERNSHIPS PROJECT

Project aims and objectives

Tools and software used

Physiological sensors: BVP and SC

Russell's circumplex model of emotions.

8

PROJECT OBJECTIVE

Design and develop a software module that

detects in real time the emotional state of the

learner.

This will be done using non intrusive physiological

sensors.

Once developed, this module is to be integrated

into an intelligent tutoring system (ITS).

9

TECHNOLOGY USED

C++ programming language

Visual Studio 2008 Integrated Development

Environment (IDE)

Thought Technology

SDK from Thought Technology software

Physiological sensors

MFC (Microsoft Foundation Classes)

10

SENSORS

Blood volume pulse (BVP) : Monitors the amount of blood perfusion in the

finger tip. Emits a small infrared light that measures the

amount of light reflected in the skins surface. The amount of blood present varies with each

heartbeat. From the BVP signal, we are able to calculate the

heart rate (HR).

11

SENSORS

Skin Conductance Sensor (SC) : Measures the skins ability to conduct electricity. Palm sides of the hands tend to become more

moist when feeling stressed or nervous. Skin with a higher sweat content will be able to

conduct an electric current more easily. SC signals are used to derive galvanic skin

response (GSR).

12

RUSSELL’S CIRCUMPLEX MODEL

HR and GSR together are used to measure specific emotions characterized in terms of valence (negative to positive) and arousal (low to high).

13

PROJECT PROGRESS

Project preparation

C++

SDK

Thought Technology

MFC

50_Read_Data

14

PROJECT PROGRESS

50_Read_Data Uses the SDKs DLL TTLLiveCtrl.dll to read in real

time.

15

PROJECT PROGRESS

Log file:

16

MFC_Connect_Encoder My first attempt of connecting to the encoder

using the SDK DLL.

PROJECT PROGRESS

17

MFC_Connect_Encoder Uses the SDK DLL TTLLiveCtrl.dll to connect to

the encoder in real time. Code example:

PROJECT PROGRESS

liEncoderHND = TTLLive->OpenConnection(L"USB:0",1000);

liEncoderCount = TTLLive->EncoderCount;if( liEncoderCount > 0 ){

m_ListDisplay.AddString("Encoder connected..");

}else {

m_ListDisplay.AddString("No encoder detected.\n\r");}

18

PROJECT PROGRESS

Connection to encoder

19

PROJECT PROGRESS

No encoder found

20

DEVELOPMENT PLAN

Read physiological signals Data processing

Heart rate (HR) Galvanic skin response (GSR)

Establishing the dashboard BVP and SC readings HR and GSR readings Russell’s model Iconic face Base line

Recording the log file

21

QUESTIONS