Emerging Topics in Computer Vision

9.3. Face Alignment

Both shape and texture (i.e., image pixels enclosed in the facial outline) provide important clues useful for characterizing the face [13]. Accurate extraction of features for the representation of faces in images offers advantages for many applications and is crucial for highly accurate face recognition and synthesis. The task of face alignment is to accurately locate facial features such as the eyes, nose, mouth, and outline, and normalize facial shape and texture.

The active shape model (ASM) [23] and active appearance model (AAM) [22, 29] are two popular models for the purpose of shape and appearance modeling and extraction. The standard ASM consists of two statistical models: (1) global shape model, which is derived from ...