Compute targets for predictions of single image. Args: gt_bboxes (Tensor): Ground truth bbox of each instance, sh

Equalizes the histogram of one image.

Slice the infinite indices by rank.

Slice the infinite indices by rank.

Infinitely yield a sequence of indices.

Infinitely yield a sequence of indices.

Initialize ``auxiliary_head``

Initialize centripetal layers. Including feature adaption deform convs (feat_adaption), deform offset prediction convs (dcn_off), gui

Initialize corner embedding layers. Only include corner embedding branch with two parts: prefix `tl_` for top-left and `br_` for bott

Initialize corner keypoint layers. Including corner heatmap branch and corner offset branch. Each branch has two parts: prefix `tl_`

Initialize the W&B Tables for validation data.

Initialize the W&B Tables for validation data.

Initialize ``decode_head``

Check and initialize input transforms. The in_channels, in_index and input_transform must match. Specifically, when input_transform i

Initialize layers of the head.

Initialize a sparse set of proposal boxes and proposal features.

Initialize layers of the head.

Initialize layers of the head.

A helper function to take a config setting and turn it into a network.

Initialize layers of the head.

Initialize layers for CornerHead. Including two parts: corner keypoint layers and corner embedding layers

Initialize layers of the transformer head.

Initialize the W&B Tables for model evaluation.

Initialize predictor layers of the head.

ViT weight initialization * When called without n, head_bias, jax_impl args it will behave exactly the same as my original init for compati

ViT weight initialization * When called without n, head_bias, jax_impl args it will behave exactly the same as my original init for compati

Lay instance masks to a result map. Args: bboxes: The bboxes results, (K, 4). labels: The labels of bboxes, (K, ).

Private function to load mask annotations. Args: results (dict): Result dict from :obj:`mmdet.CustomDataset`. Returns:

Private function to load mask and semantic segmentation annotations. In gt_semantic_seg, the foreground label is from `0` to `num_thi

Private function to load semantic segmentation annotations. Args: results (dict): Result dict from :obj:`dataset`. Retur

Load weights from .npz checkpoints for official Google Brain Flax implementation

Load weights from .npz checkpoints for official Google Brain Flax implementation

Log model checkpoint as W&B Artifact. Args: model_path (str): Path of the checkpoint to log. aliases (list): List of

Log the W&B Tables for validation data as artifact and calls `use_artifact` on it so that the evaluation table can use the reference o

Log the W&B Tables for validation data as artifact and calls `use_artifact` on it so that the evaluation table can use the reference o

Log the W&B Tables for model evaluation. The table will be logged multiple times creating new version. Use this to compare models at

Build multiple branch.

use deconv layers to upsample backbone's output.

Build fuse layer.

Make each layer.

Build one branch.

Make stem layer for ResNet.

Args: cls_pred (Tensor): Predicted classfication logits in shape (num_query, d1, ..., dn), dtype=torch.float32.

Run forward function and calculate loss for mask head in training.

Run forward function and calculate loss for mask head in training.

Run forward function and calculate loss for mask head in training.

Run forward function and calculate loss for mask head in training.

Run forward function and calculate loss for point head in training.

Export mask refining process with point head to onnx. Args: x (tuple[Tensor]): Feature maps of all scale level. rois

MixUp transform function. Args: results (dict): Result dict. Returns: dict: Updated result dict.

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

Mosaic transform function. Args: results (dict): Result dict. Returns: dict: Updated result dict.

Mosaic transform function. Args: results (dict): Result dict. Returns: dict: Updated result dict.

Mosaic transform function for label annotations. Args: results (dict): Result dict. Returns: dict: Updated r

Export the process of sampling fine grained feats to onnx. Args: x (tuple[Tensor]): Feature maps of all scale level.

Pad images according to ``self.size``.

Pad images according to ``self.size``.

Pad masks according to ``results['pad_shape']``.

Pad semantic segmentation map according to ``results['pad_shape']``.

Pad masks according to ``results['pad_shape']``.

Convert panoptic results to COCO panoptic json style.

Parse annotations and load panoptic ground truths. Args: img_info (int): Image info of an image. ann_info (list[dict]

Parse bbox and mask annotation. Args: ann_info (list[dict]): Annotation info of an image. with_mask (bool): Whether t

Private function to convert masks represented with polygon to bitmaps. Args: mask_ann (list | dict): Polygon mask annotat

Compute the area of a component of a polygon. Using the shoelace formula: https://stackoverflow.com/questions/24467972/calculate-area

Positiong embeding method. Resize the pos_embed, if the input image size doesn't match the training size. Args:

Convert proposal results to COCO json style.

Get another random index from the same group as the given index.

Ramdom jitter positive proposals for training.

Randomly sample an img_scale according to ``ratio_range`` and ``multiscale_mode``. If ``ratio_range`` is specified, a ratio will be s