Generate a CropTransform so that the cropping region contains the center of the given instance. Args: crop_size (tuple): h, w in

Computes the mode of elements in an array. Ties don't matter. Ties are broken by the smallest value (np.argmax defaults) @param arra

Computes the mode of elements in an array. Ties don't matter. Ties are broken by the smallest value (np.argmax defaults) @param arra

r""" Computes hypersphere attention on query, key and value tensors, using an optional attention mask if passed, and applying dropout if a pro

Args: image_dir (str): path to the raw dataset. e.g., "~/coco/train2017". gt_dir (str): path to the raw annotations. e.g., "~/coc

Args: image_dir (str): path to the raw dataset. e.g., "~/coco/train2017". gt_dir (str): path to the raw annotations. e.g., "~/coc

Args: image_dir (str): path to the raw dataset. e.g., "~/coco/train2017". gt_dir (str): path to the raw annotations. e.g., "~/coc

Return bbox given mask @param mask: a [H x W] numpy array

Return bbox given mask @param mask: a [H x W] torch tensor

Runs mean-shift @param X: a [n x d] torch.FloatTensor of d-dim unit vectors @param Z: a [m x d] torch.FloatTensor of seeds to run me

Runs mean-shift @param X: a [n x d] torch.FloatTensor of d-dim unit vectors @param Z: a [m x d] torch.FloatTensor of seeds to run me

More memory-friendly matching

r""" Modified from _in_projection in PyTorch.nn.functional Performs the in-projection step of the attention operation. This is simply a

r""" Modified from _in_projection_packed in PyTorch.nn.functional Performs the in-projection step of the attention operation, using packed w

Crop the image around the label mask, then resize to 224x224

Crop the image around the label mask, then resize to 224x224

Crop the image around the label mask, then resize to 224x224

Pad a possibly non-square matrix to make it square. :Parameters: matrix : list of lists matrix to pad

Parse input arguments

Parse input arguments

Parse input arguments

Parse input arguments

Parse input arguments

Parse input arguments

Parse input arguments

Parse input arguments

Convenience function: Displays the contents of a matrix of integers. :Parameters: matrix : list of lists Matrix to print

Process depth channel - change from millimeters to meters - cast to float32 data type - add random no

Process depth channel - change from millimeters to meters - cast to float32 data type - add random no

Process foreground_labels - Map the foreground_labels to {0, 1, ..., K-1} @param foreground_labels: a [H x W] numpy arra

Process foreground_labels - Map the foreground_labels to {0, 1, ..., K-1} @param foreground_labels: a [H x W] numpy arra

Process foreground_labels - Map the foreground_labels to {0, 1, ..., K-1} @param foreground_labels: a [H x W] numpy arra

Process foreground_labels - Map the foreground_labels to {0, 1, ..., K-1} @param foreground_labels: a [H x W] numpy arra

Register a "standard" version of ADE20k panoptic segmentation dataset named `name`. The dictionaries in this registered dataset follows detec

Register a "standard" version of ADE20k panoptic segmentation dataset named `name`. The dictionaries in this registered dataset follows detec

Rotate img <angle> degrees counter clockwise w.r.t. center of image @param img: a [H x W x C] image (could be an RGB image, flow image, or l