rofunc.utils.robolab.rdf.rdf#

1.  Module Contents#

1.1.  Classes#

RDF

1.2.  Functions#

job

sample_sdf_points

1.3.  API#

class rofunc.utils.robolab.rdf.rdf.RDF(args)[source]#

Initialization

Use Bernstein Polynomial to represent the SDF of the robot

train()[source]#
sdf_to_mesh(model, nbData, use_derivative=False)[source]#
create_surface_mesh(model, nbData, vis=False, save_mesh_name=None)[source]#
get_whole_body_sdf_batch(points, joint_value, model, base_trans=None, use_derivative=True, used_links=None)[source]#
get_whole_body_sdf_with_joints_grad_batch(points, joint_value, model, base_trans=None, used_links=None)[source]#

Get the SDF value and gradient of the whole body with respect to the joints

Parameters:

  • points – (batch_size, 3)

  • joint_value – (batch_size, joint_num)

  • model – the trained RDF model

  • base_trans – the transformation matrix of base pose, (1, 4, 4)

  • used_links – the links to be used, list of link names

Returns:

get_whole_body_normal_with_joints_grad_batch(points, joint_value, model, base_trans=None, used_links=None)[source]#

Get the normal vector of the whole body with respect to the joints

Parameters:

  • points – (batch_size, 3)

  • joint_value – (batch_size, joint_num)

  • model – the trained RDF model

  • base_trans – the transformation matrix of base pose, (1, 4, 4)

  • used_links – the links to be used, list of link names

Returns:

visualize_reconstructed_whole_body(model, trans_list, tag)[source]#

Visualize the reconstructed whole body

Parameters:

  • model – the trained RDF model

  • trans_list – the transformation matrices of all links

  • tag – the tag of the mesh, e.g., ‘BP_8’

Returns:

rofunc.utils.robolab.rdf.rdf.job(args)[source]#
rofunc.utils.robolab.rdf.rdf.sample_sdf_points(mf, mesh_name, save_path)[source]#