Returns torch.sqrt(torch.max(0, x)) but with a zero subgradient where x is 0.

Returns torch.sqrt(torch.max(0, x)) but with a zero subgradient where x is 0.

Get COCO API "tasks" (i.e. iou_type) from COCO-format predictions.

Process network output to extract 3D points and confidence values. Args: out: Network output tensor (B, C, H, W) activation:

Add config for common configuration

Add config for NAT Backbone.

Add config for ONE_FORMER.

Add config forSWIN Backbone.

Align `points_src` to `points_tgt` with respect to a shared xyz scale and z shift. It is similar to `align_affine` but scale and shift are a

Align `points_src` to `points_tgt` with respect to a shared xyz scale and z shift. It is similar to `align_affine` but scale and shift are a

world_tracks: 1 1 K 3 this is the coarse 3d tracks in world coordinate (coarse 3d tracks) intrs: B T 3 3 this is the intrinsic matrix

Compute the DICE loss, similar to generalized IOU for masks Args: inputs: A float tensor of arbitrary shape. The pred

Convert Batched Pytorch Tensors to PyCOLMAP Check https://github.com/colmap/pycolmap for more details about its format

Build a pixel decoder from `cfg.MODEL.MASK_FORMER.PIXEL_DECODER_NAME`.

Create a list of default :class:`Augmentation` from config. Now it includes resizing and flipping. Returns: list[Augmentation]

Build a instance embedding branch from `cfg.MODEL.INS_EMBED_HEAD.NAME`.

We estimate uncerainty as L1 distance between 0.0 and the logit prediction in 'logits' for the foreground class in `classes`. Args:

Computes the inverse of each 4x4 SE3 matrix in the batch. Args: - se3 (Tensor): Nx4x4 tensor of SE3 matrices. Returns: - Tensor

Compute the inverse of each 4x4 (or 3x4) SE3 matrix in a batch. If `R` and `T` are provided, they must correspond to the rotation and transl

Map (x, y) -> color in (R, G, B). 1) Normalize x,y to [0,1]. 2) Combine them into a single scalar c in [0,1]. 3) Use matplotlib's col

adapted from https://github.com/amyxlase/relpose-plus-plus

Compute the translation and rotation error between two camera poses. Parameters: - R1, R2: (3, 3) numpy arrays representing rotation mat

Compute the weighted standard deviation of 2D tracks across time. Args: track2d (Tensor): shape (B, T, N, 2), 2D tracked points.

Instead of storing the entire correlation pyramid, we compute each level's correlation volume, sample it immediately, then discard it

Creates a intri matrix from focal length and principal point. Args: focal_length (torch.Tensor): A Bx2 or BxSx2 tensor containing th

Create a normalized UV grid of shape (width, height, 2). The grid spans horizontally and vertically according to an aspect ratio, ensuri

Crops positional embeddings for SD3 compatibility.

r""" Decode a batch of images. Args: z (`torch.Tensor`): Input batch of latent vectors. return_dict (`bool`,

args: xy_depth: [S H W 3] focal_length: [S] R: [S 3 3] W2C T: [S 3] W2C return: xyz: [S 3 (H W)

Convert a depth map to camera coordinates. Args: depth_map (np.ndarray): Depth map of shape (H, W). intrinsic (np.ndarray):

Convert a depth map to world coordinates. Args: depth_map (np.ndarray): Depth map of shape (H, W). intrinsic (np.ndarray): C

Apply disk blur to an image using FFT convolution. Args: image (np.ndarray): Input image, can be grayscale or color. radius

Generate disk kernel with given radius. Args: radius (int): Radius of the disk (in pixels). Returns: np.ndarray

Draws keypoints of an instance and follows the rules for keypoint connections to draw lines between appropriate keypoints. This follo

Args: boxes (Boxes, RotatedBoxes or ndarray): either a :class:`Boxes`, or an Nx4 numpy array of XYXY_ABS format f

Args: circle_coord (list(int) or tuple(int)): contains the x and y coordinates of the center of the circle.

Draw instance-level prediction results on an image. Args: predictions (Instances): the output of an instance detection/se

Draw panoptic prediction annotations or results. Args: panoptic_seg (Tensor): of shape (height, width) where the values a

Draw a rotated box with label on its top-left corner. Args: rotated_box (tuple): a tuple containing (cnt_x, cnt_y, w, h,

r""" Encodes the prompt into text encoder hidden states. Args: prompt (`str` or `List[str]`, *optional*):

r""" Encodes the prompt into text encoder hidden states. Args: prompt (`str` or `List[str]`, *optional*):

r""" Encodes the prompt into text encoder hidden states. Args: prompt (`str` or `List[str]`, *optional*):

r""" Encodes the prompt into text encoder hidden states. Args: prompt (`str` or `List[str]`, *optional*):

Evaluate/summarize the performance, after processing all input/output pairs. Returns: dict: A new evalua

Args: img_ids: a list of image IDs to evaluate on. Default to None for the whole dataset

Args: img_ids: a list of image IDs to evaluate on. Default to None for the whole dataset

Convert camera extrinsics and intrinsics to a compact pose encoding. This function transforms camera parameters into a unified pose encoding form