adjoint_function

class deepinv.physics.adjoint_function(A, input_size, device='cpu', dtype=torch.float32)[source]

Bases:

Provides the adjoint function of a linear operator \(A\), i.e., \(A^{\top}\).

The generated function can be simply called as A_adjoint(y), for example:

>>> import torch
>>> from deepinv.physics.forward import adjoint_function
>>> A = lambda x: torch.roll(x, shifts=(1,1), dims=(2,3)) # shift image by one pixel
>>> x = torch.randn((4, 1, 5, 5))
>>> y = A(x)
>>> A_adjoint = adjoint_function(A, (4, 1, 5, 5))
>>> torch.allclose(A_adjoint(y), x) # we have A^T(A(x)) = x
True

Parameters:

A (callable) – linear operator \(A\).
input_size (tuple) – size of the input tensor e.g. (B, C, H, W). The first dimension, i.e. batch size, should be equal or lower than the batch size B of the input tensor to the adjoint operator.
device (str) – device where the adjoint operator is computed.

Returns:

(Callable) function that computes the adjoint of \(A\).