A Combined Feature-Texture Similarity Measure for Face Alignment under

We formulate face alignment as a model-based parameter estimation problem in this paper. First, we work within a framework that combines two separate subspace models to r epresent fr ontal face patterns and pose change independently. The combined uni ed nonlinear model represents varying pose fac es with a complex manifold. Then, we use a featur ebased similarity measure(FBSM) to evaluate image di er enc esin terms of pose, and match unknown pose faces with the model image using a combined featur e-textur e similarity measure(FTSM). Noticeable pr operties of the combine dFTSM include (1) its sensitivity to spatial differenc esbetwe enfeatur epoints in two images, which is crucial to aligning two initially faraway poses (2) easy determination of hill-climb directions in parameter space, without computing gradients of error functions. Experimental results demonstrate that, in the absence of signi cant clutter, a face alignment algorithm using the combined FTSM, can reliably align varyin...

• Type: Conference
• Year: 2000
• Where: CVPR
• Authors: Lixin Fan, Kah Kay Sung
• Keywords: CVPR 2000, Face Alignment, Match Unknown Pose, Pose Fac Es