mergeable properties for augmented tensors

[tflahaul]

As of today we can only merge tensors that have the exact same properties. Trying to merge tensors with different .projection or .distortion will raise

AssertionError: Trying to merge augmented tensor with different property: projection

Being able to merge these properties in lists would be great.

Expected return of a torch.cat((a, b, c), dim=0) with a, b & c being tensors with different projection properties:

tensor(
	projection=['kumler_bauer', 'equidistant', 'pinhole']
	normalization=minmax_sym
	pose=torch.Size([3, 4, 4])
	mask=torch.Size([3, 1, H, W])
	depth=torch.Size([3, 1, H, W])
	cam_intrinsic=torch.Size([3, 4, 4])
	cam_extrinsic=torch.Size([3, 4, 4])
	distortion=[[0.45, 0.39, 0.12], 0.25, 1.0]