Skip to main content

Advertisement

Table 6 Evaluation with different test datasets

From: A hybrid framework combining background subtraction and deep neural networks for rapid person detection

Test set Input Precision Recall Accuracy
Person Non-pers. Person Non-pers.
(a) Full-set model
 Roadset1 32 × 32 0.78 0.65 0.73 0.71 0.72
64 × 64 0.79 0.64 0.70 0.73 0.71
128 × 128 0.79 0.63 0.69 0.75 0.71
 Roadset2 32 × 32 0.84 0.80 0.80 0.84 0.82
64 × 64 0.82 0.81 0.81 0.81 0.81
128 × 128 0.83 0.79 0.78 0.84 0.80
 Nightset 32 × 32 0.76 0.50 0.72 0.55 0.66
64 × 64 0.77 0.50 0.72 0.56 0.66
128 × 128 0.76 0.48 0.68 0.58 0.65
(b) Part-set model
 Roadset1 32 × 32 0.79 0.65 0.73 0.72 0.72
64 × 64 0.81 0.66 0.73 0.75 0.74
128 × 128 0.80 0.64 0.71 0.73 0.72
 Roadset2 32 × 32 0.82 0.81 0.81 0.82 0.81
64 × 64 0.83 0.81 0.82 0.82 0.82
128 × 128 0.82 0.80 0.80 0.82 0.81
 Nightset 32 × 32 0.76 0.51 0.74 0.54 0.67
64 × 64 0.78 0.52 0.73 0.58 0.68
128 × 128 0.76 0.49 0.71 0.55 0.65