Attention Tracking Tool

Attention Tracking Tool

February 22, 2012

Review of What We are Doing• Develop an application that can track and detect

attention of students in a classroom• Generate useful metrics that can be used to

evaluate a lecture.

Milestone Four Task Matrix

TASK COMPLETION

DARYLE ABDULLAH

FINISH REPORT GENERATION 100% Write 0% Write 100%IMPLEMENT/TEST ATTENTION DETECTION

25% Write 25% Write 0%

CONTINUE EXPERIMENTING WITH FACE RECOGNITION ALGORITHMS

50% 50% 0%

POSTER FOR SENIOR DESIGN SHOWCASE

0% Write 0% Write 0%

Head Movement Tracker• Find features on the face that can be easily found in

subsequent frames.• Keep track of these points.• Use centroid of face as reference point.• Use position of features relative to reference point

to determine if person is looking left, right, up, or down.

Lessons Learned• Computer Vision is a very challenging field– Knowledge of Linear Algebra and Calculus is a must!!

• Tracking– Background subtraction (Codebooks, MOG)

• Face Recognition– Assume 1 face and high resolution image– Lots of approaches– Not always accurate but good datasets produce pretty

good results

Lessons Learned (Cont)

Demo

Questions