numpy operations:
ex. a + b applies that + operation on every pair in a and b
should work on arrays
in fastmvsnet, have this flag that says “use ground truth depth = false”
add to the eval fastmvsnet file:
use_gt_depth = True
| | | |
1m. 5m
| | | |
1m 5m 10m
change the eval fastmvsnet file:
96 is the number of hypothesis planes, change to 192 to double it
( last number before the gpu/cpu thing)
Action items: