cLPD.py π§ΒΆ
Source: LION/models/iterative_unrolled/cLPD.py
Warning
This file has not yet received a complete narrative and docstring audit. Its public source-level API is listed automatically below.
- class ConvODEFunc(layers: int, channels: list, instance_norm: bool = False)
First order ODE block
- ConvODEFunc.forward(self, t, x)
No docstring is available.
- class InitialVelocity(channels: int, instance_norm: bool = False)
Initial velocity for second order ODE
- InitialVelocity.forward(self, x0)
No docstring is available.
- class ConvSODEFunc(layers: int, channels: list, instance_norm: bool = False)
Second order ODE block
- ConvSODEFunc.forward(self, t, x)
No docstring is available.
- class ODEBlock(odefunc: nn.Module, tol: float = 0.001, adjoint: bool = False)
No docstring is available.
- ODEBlock.forward(self, x, eval_times=None, solver='rk4')
No docstring is available.
- class ContinuousDataProximal(layers: int, channels: list, tol: float = 0.001, adjoint: bool = True, do_secon_order: bool = False, solver: str = 'rk4', instance_norm: bool = False)
No docstring is available.
- ContinuousDataProximal.forward(self, x)
No docstring is available.
- class ContinuousRegProximal(layers: int, channels: list, tol: float = 0.001, adjoint: bool = False, do_secon_order: bool = False, solver: str = 'rk4', instance_norm: bool = False)
No docstring is available.
- ContinuousRegProximal.forward(self, x)
No docstring is available.
- class cLPD(geometry: ct.Geometry, model_parameters: LIONParameter = None)
Learn Primal Dual network with continuous blocks
- cLPD.default_parameters()
No docstring is available.
- cLPD.continous_LPD_paper()
No docstring is available.
- cLPD.cite(cite_format='MLA')
No docstring is available.
- cLPD.forward(self, g)
g: sinogram input