spyrit.core.recon.TikhonovMeasurementPriorDiag.wiener_denoise

TikhonovMeasurementPriorDiag.wiener_denoise(x: tensor, var: tensor) → tensor

Returns a denoised version of the input tensor using the variance prior.

This uses the attribute self.denoise_weights, which is a learnable parameter.

Inputs:

x (torch.tensor): The input tensor to be denoised.

var (torch.tensor): The variance prior.

Returns:

torch.tensor: The denoised tensor.