Test function without test time augmentation. Args: imgs (list[torch.Tensor]): List of multiple images img_metas (lis

Test function without test time augmentation. Args: imgs (list[torch.Tensor]): List of multiple images img_metas (lis

Test only det bboxes without augmentation. Args: x (tuple[Tensor]): Feature maps of all scale level. img_metas (list[

Test only det bboxes without augmentation. Args: x (tuple[Tensor]): Feature maps of all scale level. img_metas (list[

Calling its grid_anchors() method will raise NotImplementedError!

Generate grid anchors of a single level. Note: This function is usually called by method ``self.grid_priors``. Args:

Args: arguments (object): Parameters. sizes (list): List of window's sizes. gaps (list): List of window's gaps.

Args: arguments (object): Parameters. sizes (list): List of window's sizes. gaps (list): List of window's gaps.

Tests rotate anchor head loss when truth is empty and non-empty.

Tests convex regression losses. Args: loss_class (str): type of convex loss.

Test DOTA dataset. Args: angle_version (str, optional): Angle representations.

Test find latest checkpoint.

Tests fpn.

Tests gaussian regression losses. Args: loss_class (str): type of gaussian loss.

Tests gaussian regression losses v1. Args: loss_class (str): type of gaussian loss v1.

Test imshow det bboxes.

Tests kfiou regression loss.

Test reresnet backbone.

Test flip for rbboxes.

Test rotation for rbboxes.

Test rotated anchor inside flags.

Tests convex regression losses.

Test random crop for rbboxes.

Test resize for rbboxes.

Tests sam head loss when truth is empty and non-empty. Args: reassign (bool): If True, reassign samples.

Test SAR dataset. Args: angle_version (str, optional): Angle representations.

Test single stage forward (GPU). Args: cfg_file (str): config of single stage detector.

Test standard prior generator.

Test strides.

Test transforms.

Test two stage forward (GPU). Args: cfg_file (str): config of two stage detector.

Test version.

Check if detected bboxes are true positive or false positive. Args: det_bboxes (ndarray): Detected bboxes of this image, of shape (m, 6).

bool: whether the detector has a RoI head

bool: whether the detector has RPN

Convert oriented bounding box from the Pearson coordinate system to 2-D Gaussian distribution. Args: xy_stddev_pearson (torch.Tensor)

Convert oriented bounding box to 2-D Gaussian distribution. Args: xywhr (torch.Tensor): rbboxes with shape (N, 5). Returns:

Convert oriented bounding box to 2-D Gaussian distribution. Args: xywhr (torch.Tensor): rbboxes with shape (N, 5). Returns: