3D landmarking in multiexpression face analysis: a preliminary study on eyebrows and mouth

Area of Science:

• Medical Imaging
• Computer Vision
• Differential Geometry

Background:

• Three-dimensional (3D) facial analysis and landmarking algorithms are increasingly vital in maxillofacial surgery and medical diagnostics for identifying facial anomalies.
• Previous work established a geometric approach using differential geometry to extract 3D facial landmarks.
• The current study builds upon this foundation to address limitations in facial expression analysis.

Purpose of the Study:

• To extend a previously developed 3D facial landmarking algorithm.
• To enable automatic extraction of landmarks for eyebrows and mouth regions.
• To ensure the algorithm's robustness across different facial poses and mesh types.

Main Methods:

• The enhanced algorithm utilizes differential geometry principles, including fundamental forms and curvature analysis.
• It processes both existing and new 3D mesh data.
• The algorithm was validated on a private face database and the public Bosphorus 3D database.

Main Results:

• The extended algorithm successfully extracts landmarks from the mouth and eyebrow regions.
• The landmark extraction process is demonstrated to be pose-independent.
• The algorithm shows effectiveness on diverse 3D facial datasets.

Conclusions:

• The enhanced 3D facial landmarking algorithm provides accurate and pose-independent extraction of key facial points.
• This advancement holds significant potential for improving diagnostic accuracy and treatment planning in maxillofacial surgery.
• The ability to analyze facial expressions through reliable landmarking opens new avenues for medical research.