MIR.models.HyperVFA

class MIR.models.HyperVFA(*args, **kwargs)[source]

Hyperparameter-conditioned VFA model.

Parameters:

  • configs – VFA configuration object.

  • device – Device to run the model on.

  • return_orginal – If True, return composed grids and stats.

  • return_all_flows – If True, return flows for all decoder levels.

Forward inputs:

  • sample – Tuple (mov, fix) tensors.

  • hyper_val – Hyperparameter tensor of shape [1, 1].

Forward outputs:

Flow(s) depending on flags.

__init__(configs, device, return_orginal=False, return_all_flows=False)[source]

Methods

__init__(configs, device[, return_orginal, ...])

forward(sample, hyper_val)

Run forward registration.