RED

class deepinv.optim.RED(denoiser, *args, **kwargs)

Bases: Prior

Regularization-by-Denoising (RED) prior \(\nabla \reg{x} = x - \operatorname{D}_{\sigma}(x)\).

Parameters:

denoiser (callable) – Denoiser \(\operatorname{D}_{\sigma}\).

grad(x, sigma_denoiser, *args, **kwargs)

Calculates the gradient of the prior term \(\regname\) at \(x\). By default, the gradient is computed using automatic differentiation.

Parameters:

x (torch.Tensor) – Variable \(x\) at which the gradient is computed.

Returns:

(torch.Tensor) gradient \(\nabla_x g\), computed in \(x\).

Examples using RED:

Regularization by Denoising (RED) for Super-Resolution.

Regularization by Denoising (RED) for Super-Resolution.