Get the length of the whole dataset. Returns: int: The length of filtered dataset.

str: a "nice" summary string describing this module

Set attribute, i.e. self.name = value This reloading prevent the teacher model from being registered as a nn.Module. The teacher modu

Decode regression results (delta_x, delta_x, w, h) to bboxes (tl_x, tl_y, br_x, br_y). Args: priors (Tensor): Center proi

Box head forward function used in both training and testing. Returns all regression, classification results and a intermediate feature.

Box head forward function used in both training and testing. Args: x (tuple[Tensor]): List of multi-level img features.

Box head forward function used in both training and testing. Args: stage (int): The current stage in Cascade RoI Head.

bbox post-processing method. The boxes would be rescaled to the original image scale and do the nms operation. Args:

Initialize conv layers of a single level head.

Check the exists of data files.

Check whether the `num_classes` in head matches the length of `classes` in `dataset.metainfo`. Args: runner (:obj:`Runner

Check the integrity of data files.

Perform text cleaning before tokenizer.

Convert inputs type to support Detectron2's model. Args: batch_inputs (Tensor): Input images of shape (N, C, H, W).

Generate heatmaps of classification branch based on weighted distance map.

Crop image with a given center and size, then paste the cropped image to a blank image with two centers align. This function is equiv

Transform outputs into detections raw bbox prediction. Args: tl_heat (Tensor): Top-left corner heatmap for current level with

Decode init_proposal_bboxes according to the size of images and expand dimension of init_proposal_features to batch_size. Args:

do the switch aug process.

Draw instances of GT or prediction. Args: image (np.ndarray): The image to draw. instances (:obj:`InstanceData`): Dat

Draw instances of GT or prediction. Args: image (np.ndarray): The image to draw. instances (:obj:`InstanceData`): Dat

Draw panoptic seg of GT or prediction. Args: image (np.ndarray): The image to draw. panoptic_seg (:obj:`PixelData`):

Visualize every ``self.interval`` samples from a data batch. Args: batch_idx (int): The index of the current batch in the val loo

Draw semantic seg of GT or prediction. Args: image (np.ndarray): The image to draw. sem_seg (:obj:`PixelData`): Data

Extract features from the single modality. Args: inputs (Union[torch.Tensor, dict]): A batch of inputs. For image

Set a fake roi head config.

format various label to List[indices].

format various label to List[indices].

Network forward process. Usually includes backbone, neck and head forward without any post-processing. Args: batch_input

Forward for head part which is called after every decoder layer. Args: decoder_out (Tensor): in shape (batch_size, num_queries, c

Forward for head part which is called after every decoder layer. Args: decoder_out (Tensor): in shape (num_queries, batch_size, c

Freeze patch_embed layer, some parameters and stages.

Fuse global context feats with roi feats. Args: roi_feats (Tensor): RoI features. glbctx_feat (Tensor): Global contex

AlignDETR: Get label targets, label weights, and bbox weights based on `t`, the weighted geometric average of the confident score and

Get final border for the target size. This function generates a ``final_border`` according to image's shape. The area between ``final

Load a configuration as a python module.

Sample fine grained feats from each level feature map and concatenate them together. Args: x (tuple[Tensor]): Feature map

Post-process the output of head. Including softmax and set ``pred_label`` of data samples.

A context which ignores CUDA OOM exception from pytorch.

Re-normalize the image.

Slice the infinite indices by rank.

Initialize layers of the head.

Initialize predictor layers of the head.

Interpolate the `source` (4D) to the shape of the `target`.

Private function to load bounding box annotations. Args: results (dict): Result dict from :obj:``mmengine.BaseDataset``.

Private function to load label annotations. Args: results (dict): Result dict from :obj:``mmengine.BaseDataset``. Return

Mask head forward function used in both training and testing. Args: stage (int): The current stage in Cascade RoI Head.

Mask head forward function used in both training and testing. Args: stage (int): The current stage in Cascade RoI Head.

Generate mask logits from mask features with dynamic convs. Args: mask_feat (Tensor): Mask prototype features. Ha

Calculate global coordinate of mosaic image and local coordinate of cropped sub-image. Args: loc (str): Index for the sub

parse instance information. Args: raw_ann_info (ElementTree): ElementTree object. minus_one (bool): Whether to subtra

Parse panoptic segmentation predictions. Args: pred (dict): Panoptic segmentation predictions. img_id (int): Image id

perform precise bn.

Transform a single image's features extracted from the head into mask results. Args: kernel_preds (Tensor): Dynamic kerne

Transform a single image's features extracted from the head into mask results. Args: cls_scores (Tensor): Classification

Transform a single image's features extracted from the head into mask results. Args: cls_scores (Tensor): Classification