Args: norm_type: str: either in [softmax, sigmoid, divide_sum], heatmap: TENSOR (BATCH, C, ...) Returns: TENSOR (BAT

Apply orthographic camera projection on a 3D point Args: points3d (np.ndarray): shape (N, 3) ortho_cam (np.ndarray): shape (3, 3)

Calculate projective transformation of vertices given a projection matrix Input parameters: K: batch_size * 3 * 3 intrinsic c

Calculate ortho transformation of vertices given a projection parameters Input parameters: s: batch_size * 1 * 2 ; scale

Return a simplified version of this mesh. A Qslim-style approach is used here. :param factor: fraction of the original vertices to retain

Given a quaternion representing rotation, get the quaternion representing its inverse. Args: quaternion: Quaternions as tensor of

Multiply two quaternions representing rotations, returning the quaternion representing their composition, i.e. the versor with nonnegative re

Convert quaternions to axis angles. Args: quaternions (Union[torch.Tensor, numpy.ndarray]): input shape should be (..., 3

Convert rotations given as quaternions to axis/angle. Args: quat: quaternions with real part first, as tensor of shape (.

Adaptive 2D-1D registration :param vertex: 3D mesh xyz :param uv: 2D pose :param j_regressor: matrix for vertex -> joint :param K

Constructs a ResNet-18 model. Args: pretrained (bool): If True, returns a model pre-trained on ImageNet

Constructs a ResNet-34 model. Args: pretrained (bool): If True, returns a model pre-trained on ImageNet

Constructs a ResNet-50 model. Args: pretrained (bool): If True, returns a model pre-trained on ImageNet

Convert rotation 6d representations to axis angles. Args: rotation_6d (Union[torch.Tensor, numpy.ndarray]): input shape s

Converts 6D rotation representation by Zhou et al. [1] to rotation matrix using Gram--Schmidt orthogonalization per Section B of [1]. Arg

Convert rotations given as rotation matrices to Euler angles in radians. Args: matrix: Rotation matrices as tensor of shape (..., 3,

Convert rotation matrixs to quaternions. Args: matrix (Union[torch.Tensor, numpy.ndarray]): input shape should be (..., 3

Converts rotation matrices to 6D rotation representation by Zhou et al. [1] by dropping the last row. Note that 6D representation is not uniq

Convert rotations given as axis/angle to quaternions. Args: rotvec: Rotations given as a vector in axis angle form, as a

Convert rotations given as axis/angle to quaternions. Args: rotvec: Rotations given as a vector in axis angle form, as a

Convert scipy sparse matrices to pytorch sparse matrix.

Setup distributed data parallel Args: arg (Namespace): arguments rank (int): rank of current process world_size (int): to

Setup distributed data parallel Args: arg (Namespace): arguments rank (int): rank of current process world_size (int): to

Setup distributed data parallel Args: arg (Namespace): arguments rank (int): rank of current process world_size (int): to

Split scope and key. The first scope will be split from key. Examples: >>> Registry.split_scope_key('mmdet.ResNet')

Triangulate multiple 2D points from one set of multiviews using the DLT algorithm. Args: pts (np.ndarray): 2D points in the image plane. S

Triangulate one point from one set of multiviews using the DLT algorithm. Implements a linear triangulation method to find a 3D point. For exa

Computes a quadric for each vertex in the Mesh. Returns: v_quadrics: an (N x 4 x 4) array, where N is # vertices.

view a falttened 2D heatmap to 3D heatmap, sharing the same memory by using view() Args: x: TENSOR (BATCH, NCLASSES * DEP

cam is 3D [fx, fy, px, py]

Index an element from the dataset. This is done by randomly choosing a dataset using the mixing percentages and then randomly