PhD Research Journals

Found a brief FACS with sample images at http://www.cs.cmu.edu/afs/cs/project/face/www/facs.htm

Before you understand FACS, it is important to know the basic face features.


GLABELLA---> Area of the forehead between the eyebrows.
ROOT OF NOSE---> The beginning of the nose between the eyes; also called the nasal root.
EYE APERTURE---> The degree to which the eye is open; the eye opening.
EYE COVER FOLD---> The skin between the eyebrows and the palprebral part of the upper eyelid (the part that contacts the eyeball), which folds into the eye socket.
LOWER EYELID FURROW ---> A place below the lower eyelid where a line or wrinkle may appear. A line or wrinkle may be permanently etched into the face; if so, it will deepen with certain AUs. If not, it should appear when these AUs are contracted.

INFRAORBITAL FURROW ---> A place where a line or wrinkle may appear parallel to and below the lower eyelid running from near the inner corner of the eye and following the cheek bone laterally.

NOSTRIL WINGS ---> The fleshy skin of the side of the nose that forms the outside of each nostril.

NASOLABIAL FURROW ---> A place where a line or wrinkle may appear which begins adjacent to the nostril wings and runs down and outwards beyond the lip corners. In some people it is permanently etched in the face; if so, it will deepen with certain AUs. If not, it will appear on most peoples' faces with certain AUs.

PHILTRUM ---> The vertical depression in the center of the upper lip directly under the tip of the nose.

CHIN BOSS ---> The skin covering the bone of the chin.
SCLERA ---> The white part of the eyeball.

Reference: FACS Manuals

There so many acronyms available but I reckon not a single soul would even remember or know half of it. The common used ones are e.g., et al. (yes, you will only get to see these in technical reports and journals), SARS, AIDS, DJ, KL, KQ...w.r.t.??? Gosh! I am not sure if I continue using acronyms, will it means that I am getting lazier?

A note to self reminder:
w.r.t. = with respect to

My supervisor wanted me to implement an automatic landmarking, however as I was reading and reviewing on landmarks, I found a paper on automatic landmarking which is based on stereo method onto 3D object. The thought of the idea was indeed brilliant, however the more I read, I realised that setting your landmarks manually seems to be a better option. This is to enable in creating an accurate model and any errors on correspondence points can be reduced. Not forgetting, in implementing landmark, the number of landmarks is also important.

The few questions to ponder and answer:
1. Study and compare the accuracy and the effectiveness of automatic landmarking and manual landmarking.
2. Study and identify the framework for landmarking faces.
3. Identify how many landmarks were used and justify the reasons to the choice.
4. Identify the step-by-step method in implementing landmarking and the next process.

Automatic landmarking
Author: A Naftel
Title: Visualisation of Shape Change by Stereo-assisted landmark detection: an application to medical imaging (2003)
Abstract:
Techniques for the 3-D analysis of facial shape changes resulting from growth or surgical treatment have recently been proposed. These approaches rely on manual intervention for placing anatomical landmarks on 3-D facial models which is time consuming and error prone. In this paper, an automated approach to landmark extraction is described which uses a stereo-assisted active shape model. A landmark-based statistical analysis of facial shape change is then carried out using procrustes registration, principal component analysis and thin plate spline warping. The proposed method is validated both statistically and visually by characterizing shape changes induced by mandibular repositioning in a cross-sample of 20 orthodontic patients. It is shown that the method is capable of distinguishing between changes in facial morphology due to mandibular repositioning and changes due to other factors such as growth and normal variation within the patient sample.


Authors: Nafterl, Mao, Trenouth
Title: Stereo-assisted landmark detection for the anaysis of 3D facial shape changes, technical report (2002)

I've managed to install and build VtkQt. It took me more than 2 weeks to do it. I followed the instruction step-by-step however, problems faced were mainly due to the management of my Linux system. Managed to solve it, thanks to the Internet -newsgroup - and Raj who has help me much in many ways. Praise God that VtkQt works, thus, I went straight to working on retrieving vtk faces (without texture) and implementing manual landmarking.

Note regarding VtkQt, there are a few bugs. The main one that concern me was the font. The fonts are being replaced with square boxes. Sigh...Sometime I wonder if VtkQt will benefit me despite the trouble and time I have spent for nothing. Presently, my work does not even requires Qt.
....speechless......