spyrit.core.noise.NoNoise.forward

NoNoise.forward(x: tensor) → tensor

Simulates measurements

Args:

x: Batch of images. The input is directly passed to the measurement operator, so its shape depends on the type of the measurement operator.

Output:

The batch of measurements. Its shape depends on the input shape.

Shape:

x: \((*, h, w)\) if self.meas_op is a static measurement operator, \((*, t, c, h, w)\) if it is a dynamic measurement operator. Output: \((*, M)\) (static measurements) or (*, c, M) (dynamic measurements)

Example 1: Using a Linear measurement operator

>>> x = torch.FloatTensor(10, 3, 32, 32).uniform_(-1, 1)
>>> linear_acq = NoNoise(linear_op)
>>> y = linear_acq(x)
>>> print(y.shape)
torch.Size([10, 3, 400])

Example 2: Using a DynamicLinear measurement operator

>>> x = torch.FloatTensor(10, 400, 3, 32, 32).uniform_(-1, 1)
>>> dyn_acq = DynamicLinear(torch.rand(400, 32*32))
>>> noise_acq = NoNoise(dyn_acq)
>>> y = split_acq(x)
>>> print(y.shape)
torch.Size([10, 3, 400])