Complexity / 2018 / Article / Tab 21 / Research Article
An Improved EMD-Based Dissimilarity Metric for Unsupervised Linear Subspace Learning Table 21 The recognition rates on slight translation and slight rotation Extended Yale B face databases over “F1” and “F2” features; recognition rates (dimensionality).
Face databases Algorithms G6/P10 G7/P9 G8/P8 G9/P7 G10/P6 Slight translation Baseline (F1) 0.1580 (32256) 0.1622 (32256) 0.1690 (32256) 0.1783 (32256) 0.1811 (32256) LPP (F1) 0.3887 (90) 0.4339 (105) 0.4621 (105) 0.4681 (120) 0.4956 (120) Algorithm 1 (F1) 0.4758 (90)0.5035 (95)0.5196 (110)0.5419 (120)0.5544 (125)Slight translation Baseline (F2) 0.2887 (360) 0.2920 (360) 0.3052 (360) 0.3295 (360)0.3314 (360) LPP (F2) 0.2427 (125) 0.2541 (140) 0.2606 (150) 0.2726 (160) 0.2883 (170) Algorithm 1 (F2) 0.2888 (125)0.3004 (140)0.3102 (150) 0.3236 (160) 0.3311 (160) Slight rotation Baseline (F1) 0.1590 (32256) 0.1628 (32256) 0.1723 (32256) 0.1788 (32256) 0.1819 (32256) LPP (F1) 0.4005 (90) 0.4383 (105) 0.4608 (110) 0.4788 (120) 0.5019 (125) Algorithm 1 (F1) 0.5080 (90)0.5346 (100)0.5452 (105)0.5655 (110)0.5764 (120)Slight rotation Baseline (F2) 0.2833 (360) 0.2856 (360) 0.2979 (360)0.3214 (360)0.3250 (360) LPP (F2) 0.2318 (125) 0.2465 (140) 0.2494 (140) 0.2557 (160) 0.2722 (170) Algorithm 1 (F2) 0.2848 (125)0.2981 (135) 0.2967 (145) 0.3055 (140) 0.3175 (170)