Recent demos Human action and activity recognition and tracking    Mulitple human tracking from distance Real time facial feature tracking under large pose and scale changes Real time facial feature and 3D head pose tracking Worm tracking demos Multi-view face detection and tracking under significant head and facial expression changes Facial Action Units Recognition Demos

Real Time Facial Expression Recognition Under Varying Poses

Real Time Face and Eye Detection and Tracking

Facial Expression, Head Movement, Eye Movement Tracking, and Animation (joint work with Dr. Lijun Yin from SUNY Binghamton).

Facial Expression tracking and animation

Multiple Human Tracking with Particle Filtering Demos

People and face Tracking Demo

Multiple-View and multipe-people face and face pose Tracking (Indoor)

Human Tracking Demo(Outdoor)

Real Time Facial Expression Analysis demos

Real Time Facial Expression Analysis

Facial Features tracking demos

Real time facial features tracking 1

Real time facial features tracking 2

Eye gaze tracking demos

A person's gaze can provide the information where a person is looking at and what he/she is paying attention to. In the following demo, we show the possibility to control the computer cursor using the eye gaze.

map control using eye gaze

Real time eye gaze tracking We divide the screen into 4*4 grids. We use the eye gaze to control the curosr to move among these grids. A main movement is that the cursor is moved from the top-left corner to the bottom-right corner of the screen row by row.

Eye tracking demos

Real time multi-module eye tracking 1 This demo shows robustness of eye tracker after we combine the conventional kalman filter based bright pupil tracker with the recent mean shift eye tracker. This technology can robustly track the eyes under different face orientations, external strong illuminations and closed eyes.

Real time multi-module eye tracking 2

Real time multi-eye tracking This demo shows robustness of our eye tracker to detect and track eyes of different persons at the same time.

Face pose tracking demos

Real time face pose tracking 1

Real time face pose tracking 2

Real time face pose tracking 3 This demo shows robustness of our face pose tracker to detect and track face. At the same time, we do the real time face pose estimation to tell the person's face orientation. We divide the face pose into 4 types: left, right, front, up.

Real time face pose tracking 4 This demo shows robustness of our face pose tracker to detect and track face of different person. At the same time, we do the real time face pose estimation to tell the person's face orientation. We divide the face pose into 4 types: left, right, front, up.

Eyelid movement parameters tracking demos

Eyelid movement is one of the visual behaviors refect a person's level of fatigue. There are several ocular measures to characterize eyelid movement such as eye blink frequency, eye closure duration, eye closure speed. We compute two eyelid movement parameters: PERCLOS and AECS in real time and output the two parameters to computer screen. If PERCLOS and AECS measures both exceed their respective thresholds, we issue a warning alarm.
PERCLOS measures percentage of eye closure over time, exclunding the time spent on normal closure. AECS is the average eye closure speed, which means the amount of time needed to fully close the eyes and to fully open the eyes. The eye closure speed is distinctively different for a drowsy and alert subject.

Real time eyelid movement parameters display 1 In the screen, the upper one is the parameter of AECS and the lower one is the PERCLOS. The demo shows that when the person closes his eyes slowly, the AECS indicator will go up; When the person closes his eyes quickly, the AECS indicator will go down. If the person closes his eyes for a while, the PERCLOS indicator will go up; If the person opens his eyes for a while, the PERCLOS indicator will go down again.

Real time eyelid movement parameters display 2 In the screen, the upper one is the parameter of AECS and the lower one is the PERCLOS. The demo shows that when the person closes his eyes slowly, the AECS indicator will go up; When the person closes his eyes quickly, the AECS indicator will go down. If the person closes his eyes for a while, the PERCLOS indicator will go up; If the person opens his eyes for a while, the PERCLOS indicator will go down again.

Older demos

Eye tracking demos

Real time pupil tracking

Real time pupil tracking with and without glasses

Real time pupils tracking with wide angle camera the subject's face is within the view of the camera

Real time eye tracking under strong external illumnation interference

Real time eye tracking under interference of another IR light

Eye blink demos

Real time eye blink demo each time eye blinks, you hear a beep.

Eye switch demos

Real time eye switch demo the subject can turn the beep on/off by blinking three times.

Face Pose demos

Left and right face pose tracking The yellow vector represents the estimated face orientation

Up and down face pose tracking The yellow vector represents the estimated face orientation

Gaze Tracking demo

Gaze Tracking demo1 here you will see gaze (represented by a white square) follows a red laser pointer.

Gaze Tracking demo2 the same as example 1, except for the movement of the pointer is faster and larger.

Head orientation demo

		Head orientation detection demo the beeper starts beeping as soon as the subject's head rotates to the right or left.