Virtual Subspace Method for Robust Face Recognition Independent of Lighting Conditions

A concept of virtual subspace is introduced for realizing a robust face recognition independent of the lighting conditions. The virtual subspace is a paradoxical concept because it can be constructed even if only one image is taken. Furthermore, the virtual subspace is gradually converged to the real subspace when face images are subsequently taken. The virtual subspace is defined as an eigenspace composed from a synthesized image set which are supposed to be taken in a variety of lighting conditions. -4n integration algorithm is also proposed for updating the virtual subspace when additional images are available. In the experiments, we show the effectiveness of the virtual subspace method in comparison with both the conventional subspace method and the nearest neighbor discrimination.

• Type: Conference
• Year: 2000
• Where: MVA
• Authors: Takeshi Shakunaga, Nobumasa Yamamoto
• Keywords: MVA 2000, MVA 2007, Real Subspace, Subspace Method, Virtual Subspace