Do one epoch for each loader.

2020.12.24 Added to deal with attention masks

Assumes [x, y, w, h]

Transform the box co-ordinates from the original image co-ordinates to the co-ordinates of the cropped image args: box - the box for which

we should use 1 rather than image_shape

Must be deterministic

Must be deterministic

A batch collator that does nothing.

Update an attribute of the TemplateImage object with the given value. For example, `update_template("features", value)` means `self.

This function is used to update tracking information as the model runs through the video. Typical updates can be: update the local tr

r"""Wide ResNet-101-2 model from `"Wide Residual Networks" <https://arxiv.org/pdf/1605.07146.pdf>`_ The model is the same as ResNet except for

r"""Wide ResNet-50-2 model from `"Wide Residual Networks" <https://arxiv.org/pdf/1605.07146.pdf>`_ The model is the same as ResNet except for

Compute intersection over union btw segments Parameters ---------- target_segments : ndarray 2-dim array in format [m x 2:=[init,