Skip to content
Print Page Email Page PDF Version of Page

Facial Animation Laboratory


Nayoung Lee, PhD

Nayoung Lee, PhD


Current Research


Analyses of facial soft tissue shape and movement change is based on landmark changes during different facial animation/movements. Generally, landmark identification is performed manually; however, manual identification is inconvenient and requires much time and effort. Therefore, the aim of this research was to develop software to automatically identify landmarks on the circumoral region.

Fig. 1 below shows the overall flow of the proposed method of facial landmark extraction in the circumoral region for patients with cleft lip.


Fig. 1. Overall flow of the system configuration

Landmarks were positioned on the lip vermillion and circumoral region of participants as shown in Fig. 2. A horizontal center line, defined as a line through the left and right commissures (mouth corners), was identified. It was evenly divided into n+1 area by n points equidistant between the left and right commissures. Then, n points on the upper lip and n points on the lower lip, orthogonal to each of the n points on the horizontal center line, were defined as landmarks. Landmarks on the upper circumoral region then were defined using points d orthogonally distant from landmarks on the upper lip and using points 2d orthogonally distant from the landmarks along the lower lip.


Fig. 2. Landmarks in the circumoral region

As shown in Fig. 3, I have successfully developed software to automatically register landmarks in the circumoral region.


Fig. 3. Result of landmark detection in the circumoral region on 3D image

The newly-developed software can be used for comparative analyses of shape and movement between patients with CL/P and non-cleft subjects. The study sample consisted of participants recruited from the University of North Carolina School of Dentistry Orthodontic, Pedodontic, and Craniofacial Clinics. The participants were patients with repaired cleft lip and palate (CL/P) and non-cleft 'control' subjects. Supported by: NIDCR GRANT #DE013814 & #DE019742. My future work will be focused on the development of a robust and automatic facial landmark extraction.

Publications


To date, I have a total of 35 publications including the following 2 abstracts in the Facial Animation Lab:

Lee, N., Heaston, C., Rey, A., Hartman, T., Trotman, C.-A. Facial Landmark Extraction for Lip Tracking of Patients with Cleft Lip Using Active Appearance Model. Human-Computer Interaction (23). 174(4):350-354, 2011.

Lee, N., Heaston, C., Trotman, C.-A. Automatic design of landmarks in the circumoral region for patients with cleft lip. Dental Research in Review Day. University of North Carolina at Chapel Hill, 2012.