spyrit.core.meas.DynamicHadamSplit.forward_H_dyn

DynamicHadamSplit.forward_H_dyn(x: tensor) → tensor

Simulates the acquisition of measurements using the dynamic measurement matrix H_dyn.

This supposes the dynamic measurement matrix H_dyn has been set using the method build_H_dyn(). An error will be raised if H_dyn has not been set yet.

Args:

x (torch.tensor): still image of shape (*, h, w). * denotes any dimension. h and w are the height and width of the image. If h and w are larger than the measurement pattern, the image is center-cropped to the measurement pattern size.

Returns:

torch.tensor: Measurement of the input image. It has shape (*, M).