Skip to main content

Table 3 Prediction accuracy test results for drain, trash, muddy water (M.Water), water point (W.Point), water bucket (W.Bucket), tires, and animals

From: Exploring convolutional neural networks and spatial video for on-the-ground mapping in informal settlements

Video# Drain Trash M.Water     
T R1% R2% R3% T R1% R2% R3% T R1% R2% R3%     
1 15 100 100 100 10 100 100 100 9 78 89 89     
2 20 95 95 100 19 74 84 100 62 65 77 92     
3 13 100 100 100 25 76 84 100 38 74 89 95     
4 15 100 100 100 11 64 73 100 6 83 83 100     
5 6 50 100 100 15 27 53 93 8 38 38 75     
6 7 86 86 100 40 75 85 93 65 83 95 100     
7 21 100 100 100 21 57 86 90 31 55 81 97     
8 6 100 100 100 7 100 100 100 5 80 100 100     
9 4 75 75 75 16 88 94 100 11 45 73 100     
10 5 80 100 100 43 72 84 98 31 55 71 87     
11 10 100 100 100 17 76 88 94 13 69 85 100     
12 13 62 92 92 15 80 80 93 29 72 72 90     
Totals 135 91 97 99 239 72 84 96 308 68 82 94     
W. Point W. Bucket Tire Animal
T R1% R2% R3% R4% T R1% R2% R3% T R1% R2% R3% T R1% R2% R3%
4 100 100 100 100 16 88 100 100 1 100 100 100 5 100 100 100
6 50 50 83 83 27 96 96 96 18 83 89 89 7 86 100 100
3 100 100 100 100 11 100 100 100 2 100 100 100 3 67 67 67
7 86 86 86 86 9 100 100 100 1 0 100 100 1 100 100 100
5 60 60 60 80 9 78 78 78 14 93 93 93 7 86 100 100
3 33 33 67 67 4 100 100 100 2 50 50 50 8 100 100 100
4 25 25 100 100 9 100 100 100 25 88 92 92 6 67 83 83
1 100 100 100 100 1 100 100 100 1 0 0 0 1 0 0 0
3 33 33 33 100 2 100 100 100 4 100 100 100 2 100 100 100
5 80 80 80 100 2 100 100 100 3 100 100 100 3 67 100 100
2 100 100 100 100 3 100 100 100      1 0 0 0
8 100 100 100 100 12 100 100 100 7 86 100 100     
51 73 73 84 92 105 95 97 97 71 86 91 91 44 82 91 91