Skip to main content

Table 1 Recognition rates of the proposed and the related shallow networks

From: A semi-supervised convolutional neural network based on subspace representation for image classification

  CIFAR-10 [73] LFW [74] NYU Depth V1 [75] ETH-80 [76] FERET [77] Learning paradigm
PCANet [41] 78.67 85.20 79.58 93.96 97.25 Unsupervised
DCTNet [48] 73.29 85.33 75.17 89.35 97.32 Unsupervised
CCANet [54] 79.11 84.27 77.05 94.33 94.83 Unsupervised
CFR-ELM [62] 80.24 N.A. N.A. 95.63 N.A. Unsupervised
LDANet [41] 78.33 84.89 76.85 93.87 97.18 Supervised
DCCNet [58] 80.68 84.91 77.33 91.21 94.73 Supervised
OSNet [61] 78.81 83.69 76.59 94.06 93.07 Supervised
CKNet [63] 80.60 83.67 77.21 94.22 93.56 Supervised
DFSNet-1 80.77 84.96 80.31 94.23 96.96 Semi- supervised
DFSNet-2 80.97 85.29 80.53 94.43 97.28 Semi- supervised
DFSNet-3 81.06 85.45 80.61 94.52 97.47 Semi- supervised
DFSNet-4 81.20 85.55 80.68 94.66 97.54 Semi- supervised
  1. N.A. not available