Refactor load/util, start fixing packnet to support NHWC format

unsupervised/warp.py

@@ -5,6 +5,10 @@ def projective_inverse_warp(target_img, source_img, depth, pose, intrinsics):
     SFM Learner inverse warp step
         ps ~ K.T(t->s).Dt(pt).K^-1.pt
 
+    Idea is to map the pixel coordinates of the target image to 3d space (Dt(pt).K^-1.pt), then map these onto
+    the source image in pixel coordinates (K.T(t->s).{3d coord}), then using the projected coordinates we sample 
+    the pixels in the source image (ps) to reconstruct the target image.
+
     :param target_img: Tensor (batch, height, width, 3)
     :param source_img: Tensor, same shape as target_img
     :param depth: Tensor, (batch, height, width, 1)