Get the transformation matrix to align the reconstructed mesh to the ground truth mesh.

Function to compute depth and silhouette for each gaussian. These are evaluated at gaussian center.

Return paths to color images, depth images. Implement in subclass.

Calculate lr decay rate for different ViT blocks. Args: name (string): parameter name. lr_decay_rate (float): base lr decay r

r"""Inverts a 4 x 4 3D transformation matrix. Args: trans (torch.Tensor): transformation matrix (shape: :math:`* \times 4 \ti

Returns: True if the current process is the main one.

Select overlapping keyframes to the current camera observation. Args: gt_depth (tensor): ground truth depth image of the

Load camera poses. Implement in subclass.

Convert rotations given as rotation matrices to quaternions. Args: matrix: Rotation matrices as tensor of shape (..., 3, 3). Re

r"""Normalize a quaternion. The quaternion should be in (x, y, z, w) format. Args: quaternion (torch.Tensor): Quaternion to be normal

Pairwise nearest neighbors for L2-normalized vectors. Uses Torch rather than Faiss to remain on GPU.

r"""Transforms points from the pixel frame to the camera frame. Args: depth (torch.Tensor): the source depth maps (shape: :ma

Read depth data from EXR image file. Args: filename (str): File path. Returns: Y (numpy.array): Depth buffer in float32

Read embedding from file and process it. To be implemented in subclass for each dataset separately.

Resets the internal evaluation result to initial state.

Set the `seed` value for torch and numpy seeds. Also turns on deterministic execution for cudnn. Parameters:

Tag the last checkpoint. Args: last_filename_basename (str): the basename of the last filename.

r"""Applies a transform to a set of points. Args: pts (torch.Tensor): Points to be transformed (shape: B x N x D) (N points,

Update two config dictionaries recursively. Args: dict1 (dict): first dictionary to be updated. dict2 (dict): second diction

Args: pic (PIL Image, numpy.ndarray or torch.tensor): Image to be converted to tensor. Returns: Tensor: Conve