Get the shapes from a jit value object. Args: val (torch._C.Value): jit value object. Returns: list(int): return a list o

Return an activation function given a string

Count flops by input_tensor.numel() * input_scale + output_tensor.numel() * output_scale Args: input_scale: scale of the inp

Whether the input is an string instance. Note: This method is deprecated since python 2 is no longer supported.

Equivalent to nn.functional.interpolate, but with support for empty batch sizes. This will eventually be supported natively by PyTorch, and t

Args: affine_arg_index: index of the affine argument in inputs

Initializes the model. Parameters: backbone: torch module of the backbone to be used. See backbone.py transformer: to

str: a "nice" summary string describing this module

merge dict `a` into dict `b` (non-inplace). values in `a` will overwrite `b`. copy first to avoid inplace modification

[summary] Args: img ([type]): np.array, H,W,3 boxes ([type]): list of list(4) colorlist: list of colors. brands:

Run all_gather on arbitrary picklable data (not necessarily tensors) Args: data: any picklable object Returns: list[data]

Forward function.

Input: - memory: bs, \sum{hw}, d_model - memory_padding_mask: bs, \sum{hw} - spatial_shapes: nlevel, 2 - learnedw

Generalized IoU from https://giou.stanford.edu/ The boxes should be in [x0, y0, x1, y1] format Returns a [N, M] pairwise matrix, where

Args: x (float): Input images(BxCxHxW) axis (int): The index for weighted mean other_axis (int): The other index Re

label: Tensor(K)

Merge list into cfg_dict Merge the dict parsed by MultipleKVAction into this cfg. Examples: >>> options = {'model.backbo

r""" Args: query, key, value: map a query and a set of key-value pairs to an output. See "Attention Is All You Need" for more

image: A PIL image boxes: Bounding boxes, a tensor of dimensions (#objects, 4) labels: labels of object, a tensor of dimensio

Loss used in RetinaNet for dense detection: https://arxiv.org/abs/1708.02002. Args: inputs: A float tensor of arbitrary shape.

Convert the model into training mode while keep layers freezed.

return if the results is the best.

Args: x: (B, H, W, C) window_size (int): window size Returns: windows: (num_windows*B, window_size, window_size, C)

Creates the matcher Params: cost_class: This is the relative weight of the classification error in the matching cost c

str: the string of the module