Answer a question about an image Args: image: Input image question: Text question

Answer questions about a batch of images Args: images: List of input images questions: List of text

Build and initialize the VGGSfM tracker. Args: model_path: Path to the model weights file. If None, weights are downloaded from Hugg

Construct an order that switches [query_index] and [0] so that the content of query_index would be placed at [0]. Args: query_in

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

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

Count objects in an image based on a query Args: image: Input image query: Text query describing wha

Count objects in a batch of images Args: images: List of input images queries: List of text queries

Create the bounding box frame without white borders

Create the bounding box frame for depth estimation

Create the depth estimation frame

Create the original image frame for depth estimation

Create the original image frame without white borders

Create the overlay frame without white borders

Create the bounding box frame for pose estimation

Create the pose estimation frame

Create the original image frame for pose estimation

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

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

Detect objects in an image based on a query Args: image: Input image query: Text query describing wh

Extract keypoints using pre-initialized feature extractors. Args: query_image: Input image tensor (3xHxW, range [0, 1]) extr

Farthest point sampling algorithm to select diverse frames. Args: distance_matrix: Matrix of distances between frames num_sa

x_list contains a list of tensors to nest together and run

Generate an image based on a query Args: refer_image_path: Path to the reference image image_prompt:

Generate segmentation masks for given image, bounding boxes and points Args: image (PIL.Image): Input image

Generate segmentation masks for given image, bounding boxes and points Args: image (PIL.Image): Input image

Generate a ranking of frames using DINO ViT features. Args: images: Tensor of shape (S, 3, H, W) with values in range [0, 1]

This function generates a 2D positional embedding from given coordinates using sine and cosine functions. Args: - xy: The coordinates to

This function generates a 2D positional embedding from given coordinates using sine and cosine functions. Args: - xy: The coordinates to

This function initializes a grid and generates a 2D positional embedding using sine and cosine functions. It is a wrapper of get_2d_sincos_po

This function initializes a grid and generates a 2D positional embedding using sine and cosine functions. It is a wrapper of get_2d_sincos_po

This function generates a 2D positional embedding from a given grid using sine and cosine functions. Args: - embed_dim: The embedding di

This function generates a 2D positional embedding from a given grid using sine and cosine functions. Args: - embed_dim: The embedding di

Given all tracks in one sample (b), compute a (N,3) array of RGB color values in [0,255]. The color is determined by the (x,y) position in th

Applies intrinsic parameters and optional distortion to the given 3D points. Args: cam_intrinsics (torch.Tensor): Intrinsic camera p

Applies intrinsic parameters and optional distortion to the given 3D points. Args: intrinsics (torch.Tensor): Intrinsic camera param

Apply intrinsics (and optional radial distortion) to camera-space points. Args ---- intrinsics : (B,3,3) camera matrix K. point

Convert rotations given as rotation matrices to quaternions. Args: matrix: Rotation matrices as tensor of shape (..., 3, 3). Re

合并所有bboxes，返回最左上和最右下的坐标

Modulate the input tensor using scaling and shifting parameters.

Pose estimation in an image based on a query Args: image: Input image query: Text query describing w

Convert 2D position grid (HxWx2) to sinusoidal embeddings (HxWxC) Args: pos_grid: Tensor of shape (H, W, 2) containing 2D coordinate

Process a list of query points to avoid memory issues. Args: track_predictor (object): The track predictor object used for predictin

Process a batch of images and questions

Process a batch of images and questions

Process a batch of images and questions

Process a batch of images and questions

Process a batch of images and questions

Process a batch of images and questions