Research Article
Fusing Two Kinds of Virtual Samples for Small Sample Face Recognition
(1) Normalize the columns of and to have unit -norm | (2) Solve | If is singular, , where is a small | positive constant and is the identity matrix. | If is not singular, . | (3) Compute the deviation | | (4) Output the training samples that have the greatest contributions | | (5) Solve | If is singular, , where is a small | positive constant and is the identity matrix; | If is not singular, . | (6) Compute the deviation | | (7) Output the identity of as | Identity |
|