Args: feat: feature maps shaped in B * C * H * W ind: indices tensor shaped in B * K Returns:

Convert continuous angle to discrete class and residual.

:param pts: (N, 3 or 2) :return pts_hom: (N, 4 or 3)

Converts cartesian feature map to log polar :param input_feature_map: Tensor array [b x c x h x w] or numpy ar

Converts tensor or numpy array input from HWC to CHW format

top_color_frac and bottom_color_frac are colors that will be used for the two ends of the spectrum. Reference https://towardsdatascie

Computes the amount of Intersection over Union (IoU) between two different sets of boxes. Args: box_a (nparray): Mx4 boxes, defined

Converts cartesian feature map to log polar :param log_polar_feature_map: Tensor array [b x c x h x w] or numpy

Recursively make a directory structure even if missing. if delete_if_exist=True then we will delete it first which can be useful when be

Get rotation_y by alpha + theta - 180 alpha : Observation angle of object, ranging [-pi..pi] x : Object center x to the camera center (x-

Get rotation_y by alpha + theta - 180 alpha : Observation angle of object, ranging [-pi..pi] x : Object center x to the camer