Pulling frame-by-frame, pixel-by-pixel depth maps from inference script

[sunshineinsandiego]

Is there a preferred or recommended way to do this? Thanks!