Saves the output of the tracker.

Convert annotation which can be polygons, uncompressed RLE to RLE. :return: binary mask (numpy 2D array)

:param boxes1: (N, 4) (x1,y1,x2,y2) :param boxes2: (N, 4) (x1,y1,x2,y2) :return:

Eliminate potential background candidates for computation reduction and noise cancellation. Args: attn (torch.Tensor): [B, num_hea

Checks if the pre-computed results are valid

mask_encoding: a string in the following format: x0, y0, w, h, RLE output: mask, offset mask: 2-D binary mask, size defined in the mask e

Do a cycle of training or validation.

Download COCO images from mscoco.org server. :param tarDir (str): COCO results directory name imgIds (list): images to

Script to automatically download tracker results for PyTracking. args: download_path - Directory where the zipped results are downloa

Speed Test

Speed Test

cat_feature: output embeddings of the backbone, it can be (HW1+HW2, B, C) or (HW2, B, C)

Generalized IoU from https://giou.stanford.edu/ The boxes should be in [x0, y0, x1, y1] format boxes1: (N, 4) boxes2: (N, 4)

Generate ground truth heatmap same as CenterNet Args: bboxes (torch.Tensor): shape of [num_search, bs, 4] Returns:

Get img ids that satisfy given filter conditions. :param imgIds (int array) : get imgs for given ids :param catIds (int array

copyed from CornerNet box_size (w, h), it could be a torch.Tensor, numpy.ndarray, list or tuple notice: we are using a bug

Number of sequences in a dataset returns: int - number of sequences in the dataset.

Number of sequences in a dataset returns: int - number of sequences in the dataset.

Overload this in your dataset. Should return the list of sequences in the dataset.

returns: TensorDict - dict containing all the data blocks

args: index (int): Index (Ignored since we sample randomly) aux (bool): whether the current data is for auxiliary use

2D cosine window.

Load indexed image with given filename. Used to read segmentation annotations.