Skip to main content

CodedNetSDF

C++: phynexis::ml::shape::CodedNetSDF
Python: phynexis.ml.shape.CodedNetSDF
Header: src/ml/shape/coded_net_sdf.hpp

Description

Signed distance function with latent code conditioning. Extends NetSDF with a latent code that modulates the neural network output.

Constructors

CodedNetSDF()

Creates an empty coded SDF.

Methods

init_from_model_files(net_sdf_file, latent_code_file, id)

Initializes from model and latent code files.

Parameters:

ParameterTypeDescription
net_sdf_filestrPath to the SDF model file
latent_code_filestrPath to the latent code file
idintObject ID

initialize()

Initializes from memory.

signed_distance(pos)

Computes the signed distance at a position.

Parameters:

ParameterTypeDescription
posVec3dPosition

Returns: float — signed distance

Prints basic information about the coded SDF.

Inherited from NetSDF / NetField

  • init_from_model_files(model_file)
  • initialize()

Properties

PropertyTypeAccessDescription
latent_codelist[float]read/writeLatent code vector

Inherited from NetField

PropertyTypeAccessDescription
net_modelNeuralNetread/writeNeural network model
field_typeFieldTyperead/writeField type
surface_configSurfaceConfigread/writeSurface extraction configuration