Research Article

Discriminative, Competitive, and Collaborative Representation-Based Classification with -Norm Regularizations

Table 5

The classification accuracies (%) of the competing methods with the corresponding standard deviations under the situations of the random block occlusions.

DataRatiosSRCCRCCCRCProCRCEProCRCCo-CRCDSRCDCCRC

IMM0.162.17 ± 1.8063.00 ± 1.0065.17 ± 1.3365.50 ± 0.8562.33 ± 0.9763.67 ± 1.3562.67 ± 0.9766.00±0.82
0.254.33 ± 1.4356.17 ± 1.8059.50 ± 1.8758.67 ± 1.6355.50 ± 3.2755.33 ± 1.5557.50 ± 1.1859.67±1.45
0.345.67 ± 2.4447.67 ± 1.6250.00 ± 2.3650.83 ± 2.4748.17 ± 2.4449.50 ± 2.5149.83 ± 1.2251.50 ± 2.26
0.437.33 ± 3.1435.83 ± 1.9040.50 ± 2.1541.00±2.4438.33 ± 2.3639.67 ± 1.7239.33 ± 2.2041.00±2.44

Yale0.182.67 ± 2.8386.67 ± 2.1186.50 ± 2.2586.50 ± 1.9485.50 ± 2.5385.00 ± 3.1884.17 ± 3.2387.33±2.31
0.276.00 ± 2.8379.17 ± 3.4679.00 ± 1.7379.17 ± 1.7678.67 ± 1.7675.33 ± 1.1575.00 ± 1.6079.33±1.89
0.364.83 ± 3.5268.83 ± 2.4970.33 ± 2.5669.50 ± 3.0169.17 ± 2.9462.50 ± 3.2365.50 ± 3.4970.83±1.94
0.453.07 ± 2.1356.80 ± 1.3657.87 ± 2.3256.27 ± 4.4956.53 ± 4.8145.33 ± 2.5350.67 ± 2.3958.40±2.29