Nomalize each vector to unit
%===========================================
[nSmp,nFea] = size(fea);
for i = 1:nSmp
     fea(i,:) = fea(i,:) ./ max(1e-12,norm(fea(i,:)));
end
%===========================================

Scale the features (pixel values) to [0,1]
%===========================================
maxValue = max(max(fea));
fea = fea/maxValue;
%===========================================

• Deng Cai, Xiaofei He, Yuxiao Hu, Jiawei Han and Thomas Huang, "Learning a Spatially Smooth Subspace for Face Recognition", CVPR'07. Bibtex source
• Deng Cai, Xiaofei He and Jiawei Han, "Spectral Regression for Efficient Regularized Subspace Learning", ICCV'07. Bibtex source
• Deng Cai, Xiaofei He, Jiawei Han, and Hong-Jiang Zhang, "Orthogonal Laplacianfaces for Face Recognition", IEEE TIP 2006. Bibtex source
• Xiaofei He, Shuicheng Yan, Yuxiao Hu, Partha Niyogi, and Hong-Jiang Zhang, "Face Recognition Using Laplacianfaces", IEEE TPAMI 2005. Bibtex source

• Yale database

  Contains 165 grayscale images in GIF format of 15 individuals. There are 11 images per subject, one per different facial expression or configuration: center-light, w/glasses, happy, left-light, w/no glasses, normal, right-light, sad, sleepy, surprised, and wink.

  32x32 Data File: contains variables 'fea' and 'gnd'. Each row of 'fea' is a face; 'gnd' is the label. Sample images
  64x64 Data File.

  A random subset with p(=2,3,4,5,6,7,8) images per individual was taken with labels to form the training set, and the rest of the database was considered to be the testing set. For each given p, there are 50 randomly splits:
  2 Train | 3 Train | 4 Train | 5 Train | 6 Train | 7 Train | 8 Train |

  Each split file contains variables 'trainIdx' and 'testIdx'. The following matlab codes can be used to generate the training and test set:
  %===========================================
  fea_Train = fea(trainIdx,:);
  fea_Test = fea(testIdx,:);
  
  gnd_Train = gnd(trainIdx);
  gnd_Test = gnd(testIdx);
  %===========================================
  Experimental results (new!) | old experimental results

• ORL database

  Ten different images of each of 40 distinct subjects. For some subjects, the images were taken at different times, varying the lighting, facial expressions (open / closed eyes, smiling / not smiling) and facial details (glasses / no glasses). All the images were taken against a dark homogeneous background with the subjects in an upright, frontal position (with tolerance for some side movement).

  32x32 Data File: contains variables 'fea' and 'gnd'. Each row of 'fea' is a face; 'gnd' is the label. Sample images
  64x64 Data File

  A random subset with p(=2,3,4,5,6,7,8) images per individual was taken with labels to form the training set, and the rest of the database was considered to be the testing set. For each given p, there are 50 randomly splits:
  2 Train | 3 Train | 4 Train | 5 Train | 6 Train | 7 Train | 8 Train |

  Experimental results (new!) | old experimental results

• Extended Yale Face Database B

  Please refer to this link for detailed information.
  For this database, we simply use the cropped images (which can be downloaded here) and resize them to 32x32 pixels. This dataset now has 38 individuals and around 64 near frontal images under different illuminations per individual.

  32x32 Data File: contains variables 'fea' and 'gnd'. Each row of 'fea' is a face; 'gnd' is the label. Sample images
  

  A random subset with p(=5,10,...,50) images per individual was taken with labels to form the training set, and the rest of the database was considered to be the testing set. For each given p, there are 50 randomly splits:
  5 Train | 10 Train | 20 Train | 30 Train | 40 Train | 50 Train |

  Experimental results (new!) | old experimental results