Initialize ``auxiliary_head``

Initialize ``auxiliary_head``

Initialize classification branch and regression branch of head.

Initialize classification branch and regression branch of head.

Initialize ``decode_head``

Initialize ``decode_head``

Initialize layers of the transformer head.

Initialize some layer.

Initialize some layer.

Initialize classification branch and regression branch of head.

r""" Inject tp matching into predictions. Parameters ---------- num_gt : int Number of ground truth. pred : dict

r""" Calculate mAP on topology among lane centerlines. Parameters ---------- gts : dict Dict storing ground truth for all sam

r""" Calculate mAP on topology between lane centerlines and traffic elements. Parameters ---------- gts : dict Dict storing g

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

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

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

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

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

r""" Calculate average precision based on given recalls and precisions. Parameters ---------- recalls : array_like List in sh

Prepare class label and vertex context.

Prepare class label and vertex context.

Convert proposal results to COCO json style.

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

Resize images with ``results['scale']``.

Resize semantic segmentation map with ``results['scale']``.

Sample negative samples.

Sample positive samples.

Convert instance segmentation results to COCO json style.

r""" Generate lists of tp and fp on given distance threshold. Parameters ---------- gts : List List of groud truth in shape (

Test function with test time augmentation.

Calculate average precision. Args: recalls (ndarray): shape (num_dets, ) precisions (ndarray): shape (num_dets, ) mode (s

pred: B,nquery,npts label: B,nquery,npts

Matching predicted lanes and ground-truth lanes in their IPM projection, ignoring z attributes. x error, y_error, and z error

Args: mask_preds (Tensor): Mask prediction in shape (N1, H, W). gt_masks (Tensor): Ground truth in shape (N2, H, W)

Args: mask_preds (Tensor): Mask prediction in shape (N1, H, W). gt_masks (Tensor): Ground truth in shape (N2, H, W)

Build data config for loading visualization data.

Build data config for loading visualization data.

Args: logits: B1, 8, cls_dim value: B2, 8, Return: log_likelihood: B1,B2,8

Args: x1: B1,N,2 x2: B2,N,2 Return: dist_mat: B1,B2,N

Convert transformation matrix in Cartesian coordinates to homogeneous format. Args: mat (np.ndarray): Transformation matr

pred: B*nquery,npts label: B*nquery,

Check if a box is visible inside an image without accounting for occlusions. :param box: The box to be checked. :param intrinsic: <float:

Args: pred: [num_points, 2] gt: [num_gt, 2] Out: torch.FloatTensor of shape (1, )