Skip to main content

Table 2 Summary of results

From: Face recognition with Bayesian convolutional networks for robust surveillance systems

Method

Recognition accuracy (%)

EURECOM Kinect Face Database [55]

AT&T face database [56]

PCA [18]

89.0

91.0

MahCos PCA [19]

90.4

92.5

DCNN by Lee et al. [58]

97.0 [58]

DCNN by Vinay et al. [59]

95.2 [59]

Proposed DCNN

 Arc-1

90.4

87.5

 Arc-2

94.2

97.5

 Arc-3

92.3

92.5

Proposed B-DCNN

 Arc-1

94.2

92.5

Arc-2

98.1

100

 Arc-3

96.2

97.5

  1. The results of the best performing network architecture are shown in italic