Regularisation of eigenfeatures by extrapolation of scatter-matrix in face-recognition problem
Alok K. Sharma, Kuldip K. Paliwal · Electronics Letters · 2010
The proposed face recognition algorithm regularises the within-class scatter matrix by extrapolating its eigenvalues. This way both the range space and the null space information of the within-class scatter matrix is utilised. The results are promising when the algorithm is experimented on face recognition datasets and compared with several other methods.