Rescale the output instances to the target size.

Uniform float random number between low and high.

In given proto, rename all blobs with matched version

mmdet will assert the type of dict/list. So convert omegaconf objects to dict/list.

Try select keys from cfg until the first key that exists. Otherwise return default.

Accumulate instances data for one image Args: instances_one_image (Instances): instances data to accumulate

Call `add_ground_truth_to_proposals_single_image` for all images. Args: gt(Union[List[Instances], List[Boxes]): list of N elements.

Add config for PointRend.

Called after the backward pass of each iteration.

Computes area of the mask. Only works with Polygons, using the shoelace formula: https://stackoverflow.com/questions/24467972

Based on IoU for each pair of bbox, assign the associated value in cost matrix Args: cost_matrix: np.ndarray, initialize

When the config is defined for certain number of workers (according to ``cfg.SOLVER.REFERENCE_WORLD_SIZE``) that's different from the

Performs non-maximum suppression in a batched fashion. Each index value correspond to a category, and NMS will not be applied between el

Build a box head defined by `cfg.MODEL.ROI_BOX_HEAD.NAME`.

Build a point head defined by `cfg.MODEL.POINT_HEAD.NAME`.

Create a ResNet instance from config. Returns: ResNet: a :class:`ResNet` instance.

Build a semantic segmentation head from `cfg.MODEL.SEM_SEG_HEAD.NAME`.

Build a tracker head from `cfg.TRACKER_HEADS.TRACKER_NAME`. Args: cfg: D2 CfgNode, config file with tracker information Return:

See get_caffe2_inputs() below.

Convert an instance detection/segmentation or keypoint detection dataset in detectron2's standard format into COCO json format. Generic

Create a DistributedDataParallel model if there are >1 processes. Args: model: a torch.nn.Module fp16_compression: add fp16

Args: text (str): class label position (tuple): a tuple of the x and y coordinates to place text on image.

For each feature map, select the `pre_nms_topk` highest scoring proposals, apply NMS, clip proposals, and remove small boxes. Return the `pos

Use the given boxes in `instances` to produce other (non-box) per-ROI outputs. Args: features: same as in `forward()`

Run gather on arbitrary picklable data (not necessarily tensors). Args: data: any picklable object dst (int): destination ra

Generate regular square grid of points in [0, 1] x [0, 1] coordinate space. Args: R (int): The number of grids to sample, one for ea

Returns the field called `name`.

Return dict from versioned blob to list of (i, j), where i is index of consumer op, j is the index of input of that op.

Get default param list for optimizer, with support for a few types of overrides. If no overrides needed, this is equivalent to `model.paramet

Load and prepare dataset dicts for instance detection/segmentation and semantic segmentation. Args: names (str or list[str]): a data

Returns: dict: a dict which maps names (str) to data of the fields Modifying the returned dict will modify this instance

Load LVIS metadata. Args: dataset_name (str): LVIS dataset name without the split name (e.g., "lvis_v0.5"). Returns: di

Return dict from versioned blob to (i, j), where i is index of producer op, j is the index of output of that op.

Returns: Check if there are EventStorage() context existed.

Returns: dict[name -> HistoryBuffer]: the HistoryBuffer for all scalars

Return the device type ("cpu" or "gpu"/"cuda") of each (versioned) blob

Run inference on the given inputs. Args: batched_inputs (list[dict]): same as in :meth:`forward` detected_in

Prepare some proposals to be used to train the ROI heads. It performs box matching between `proposals` and `targets`, and assigns

Similar to :meth:`load()`, but load path relative to the caller's source file. This has the same functionality as a relative

Load semantic segmentation datasets. All files under "gt_root" with "gt_ext" extension are treated as ground truth annotations and all files

This method is almost identical to :meth:`RetinaNet.losses`, with an extra "loss_centerness" in the returned dict.

Returns True when torch's version is at least `min_version`.

Fake losses for fine segmentation and U/V coordinates. These are used when no suitable ground truth data was found in a batch. The lo

Add an `img_tensor` associated with `img_name`, to be shown on tensorboard. Args: img_name (str): The name of th

Args: box_scores (N, 5): boxes in corner-form and probabilities. (Note here 5 == 4 + 1, i.e., 4-dim horizontal bo

Save a jit module and load it back. Similar to the `getExportImportCopy` function in torch/testing/

Rename the op_id-th operator in predict_net, change it's output_id-th input's name to the new_name. It also does automatic re-route and c

Compute (fractional) per-image repeat factors based on category frequency. The repeat factor for an image is a function of the freque

Resample fine and coarse segmentation outputs from a predictor to the given bounding box and derive labels for each pixel of the bounding box

Compute the point-based loss for instance segmentation mask predictions given point-wise mask prediction and its corresponding point-wise lab

Turn a list of integer scalars or integer Tensor scalars into a vector, in a way that's both traceable and scriptable. In tracing, `x` s

Args: img_size (int): Input image size. patch_kernel (tuple): kernel size for patch embedding. patch_stri

Args: bottom_up (Backbone): module representing the bottom up subnetwork. Must be a subclass of :class:`Backbone`

NOTE: this interface is experimental. Args: num_classes (int): number of foreground classes (i.e. background is not incl

Args: cfg (CfgNode): torch_model (nn.Module): the detectron2 model (meta_arch) to be converted.

Args: conv_dim: the output dimension of the conv layers fc_dims: a list of N>0 integers representing the output dimen

Assert that each stride is 2x times its preceding stride, i.e. "contiguous in log2".

Naive tracking heuristics to assign same color to the same instance, will update the internal state of tracked instances. Re

For each untracked instance, assign a new id Args: instances: D2 Instances, for predictions of the current frame