about the code in p_sample_loop_known

[jaceqin]

if dice_score(final["sample"][:,-1,:,:].unsqueeze(1), final["cal"]) < 0.65:
cal_out = torch.clamp(final["cal"] + 0.25 * final["sample"][:,-1,:,:].unsqueeze(1),
else:
cal_out = torch.clamp(final["cal"] * 0.5 + 0.5 * final["sample"][:,-1,:,:].unsqueeze(1), 0, 1)

What is the main purpose of this code？

[xupinggl]

same question

[Devin-Pi]

same question.
What's more important, as show below, why the cal_out can represent the final segmentation result?
co = th.tensor(cal_out)
print('the shape of sample :', sample.shape, sample[:, -1, :, :].shape)
if args.version == 'new':
enslist.append(sample[:,-1,:,:])
else:
enslist.append(co)