Fig 1.
Graphical analysis of our proposed framework.
(a) the effect of the first constraint term and (b) the effect of the incorporated first two constraint terms.
Fig 2.
(a) Extended YaleB, (b) AR, (c) COIL20, (d) USPS.
Table 1.
Classification rates(%) of comparison methods on test datasets with KNN.
Table 2.
Classification rates(%) of comparison methods on test datasets with SRC.
Fig 3.
(a) Extended YaleB, (b) COIL20.
Fig 4.
Recognition results versus pixel corruption on extended YaleB.
Fig 5.
Classification results versus pixel corruption on COIL20.
Fig 6.
Classification results versus variational η.
Fig 7.
Classification results versus variational λ.
Fig 8.
Objective function values versus iterative steps.
(a) curve for Extended YaleB dataset and (b) curve for COIL20 dataset.