Skip to main content

Table 4 Runtime statistics for predictions at various resolutions

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

Resolution

Prediction time for single frame (s)

Max stack size

Prediction time for max stack (s)

416px

0.03

70

0.03

608px

0.02

30

0.02

832px

0.02

15

0.02

1024px

0.02

10

0.02

1280px

0.02

5

0.02