Sirface vs. Fisherface: recognition using class specific linear projection

Using a novel data dimension reduction method proposed in statistics, we develop an appearance-based face recognition algorithm which is insensitive to large variation in lighting direction and facial expression. Taking a pattern classification approach, we consider each pixel in an image as coordinate in a highdimensional space. However, since faces are not truly Lambertian surfaces and indeed produce self-shadowing, images will deviate from this linear subspace. Rather than explicitly modeling this deviation, we linearly project the image into a subspace in a manner which discounts those regions of the face with large deviation using Sliced Inverse Regression (SIR) [9]. Our face recognition algorithm termed as Sirface produces well-separated classes in a low-dimensional subspace, even under severe variation in lighting and facial expression. Sirface can be shown to be equivalent to the well known Fisherface algorithm [1] in the subspace sense. However, Sirface is shown to produce th...

• Type: Conference
• Year: 2003
• Where: ICIP
• Authors: Yangrong Ling, Xiangrong Yin, Suchendra M. Bhandarkar
• Keywords: Face Recognition Algorithm, ICIP 2003, Image Processing, Linear Subspace, Low-dimensional Subspace, Optimal Reduced Subspace, Subspace Sense