:param ctx: :param features: features of points, FloatTensor[B, C, N] :param indices: neighbor indices of centers, IntTensor[

:param ctx: :param points_coords: coordinates of points, FloatTensor[B, 3, N] :param centers_coords: coordinates of centers,

:param ctx: :param features: Features of the point cloud, FloatTensor[B, C, N] :param coords: Voxelized Coordinates of each p

:param ctx: :param coords: the coordinates of points, FloatTensor[B, 3, N] :param features: FloatTensor[B, C, R, R, R]

Gather :param ctx: :param features: features of points, FloatTensor[B, C, N] :param indices: centers' indices in poin

Forward pass of DiT, but also batches the unconditional forward pass for classifier-free guidance.

Uses iterative furthest point sampling to select a set of npoint features that have the largest minimum distance to the sampled point set

:param target_shape_constraint: target voxels :return: constrained x

Computes the JSD between two sets of point-clouds, as introduced in the paper ```Learning Representations And Generative Models For 3D Point Clouds```

Use logits to sample points :param coords: coords of points, FloatTensor[B, 3, N] :param logits: binary classification logits, FloatTenso

Save a 3D mesh to a polygon .ply file.

Computes the MMD between two sets of point-clouds. Args: sample_pcs (numpy array SxKx3): the S point-clouds, each of K points that will be

KL divergence between normal distributions parameterized by mean and log-variance.

project point clouds into 2D image :param points: (n, 3) range(-1, 1) :return: binary image

vertices: [numpoints, 3]

align 3depn shapes to shapenet coordinates

Function: trilinear devoxelization (backward) Args: grad_y : grad outputs, FloatTensor[b, c, n] indices : the voxel coordinates of point cu

Function: trilinear devoxelization (forward) Args: r : voxel resolution trainig : whether is training mode coords : the coordi

r''' Visualizes point cloud data. Args: points (tensor): point data normals (tensor): normal data (if existing) out_file (

r''' Visualizes voxel data. show only first num_shown

xavier initialization

input: Nx3, write points to filename as PLY format.