Args: - depthmap (BxHxW array): - pseudo_focal: [B,H,W] ; [B,2,H,W] or [B,1,H,W] Returns: pointmap of absolute coordi

:param flow: Optical flow tensor. :return: RGB image normalized between 0 and 1.

x_list contains a list of tensors to nest together and run

Create a composition of image augmentations. Args: color_jitter: Dictionary containing color jitter parameters: - brightn

Integrates a fake camera mesh into the 3D scene. Args: scene (trimesh.Scene): The 3D scene to add the camera model. transfor

Loads a 16-bit PNG as a half-float depth map (H, W), returning a float32 NumPy array. Implementation detail: - PIL loads 16-bit data a

Load camera information from a text file. Expected file format: - First 3 lines: 3x4 extrinsic matrix (each line has 4 space/tab-separat

Load images and corresponding camera/depth information from folders. Args: image_folder: Path to folder containing images ca

Load and initialize the OmniVGGT model. Args: cfg: Configuration object device: Target device Returns:

Main function for the OmniVGGT demo with viser for 3D visualization. This function: 1. Loads the OmniVGGT model 2. Processes input i

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

Modulate the input tensor using scaling and shifting parameters.

Normalize camera extrinsics and corresponding 3D points. This function transforms the coordinate system to be centered at the first came

Modify camera intrinsics to follow a different convention. Coordinates of the center of the top-left pixels are by default: - (0.5, 0.5)

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

Quaternion Order: XYZW or say ijkr, scalar-last Convert rotations given as quaternions to rotation matrices. Args: quaternions:

Runs sky segmentation inference using ONNX model. Args: onnx_session: ONNX runtime session input_size: Target size for model

Forward pass through the fusion blocks. Args: features (List[Tensor]): List of feature maps from different layers.

Setup output and logging directories. Args: cfg: Configuration object Returns: Tuple of (save_dir, logging_

Setup logging configuration for all processes.

Setup TensorBoard logging. Args: cfg: Configuration object save_dir: Output directory Returns: Summ

Setup Weights & Biases logging.

Simple resize of image and depthmap, adjusting intrinsics if provided. Args: image: PIL Image or numpy array depthmap: n

Read camera data, return (M,N) tuple. M is the intrinsic matrix, N is the extrinsic matrix, so that x = M*N*X, with x being a point in h

Convert a unit quaternion to a standard form: one in which the real part is non negative. Args: quaternions: Quaternions with re

Better torch.quantile for one SCALAR quantile. Using torch.kthvalue. Better than torch.quantile because: - No 2**24 input size limit (pyt

Iteratively refine camera pose predictions. Args: pose_tokens (torch.Tensor): Normalized camera tokens with shape [B, 1,

Visualize predicted 3D points and camera poses with viser. Args: pred_dict (dict): { "images": (S, 3, H,

Add camera frames and frustums to the scene. extrinsics: (S, 3, 4) images_: (S, 3, H, W)

Output a (H,W,2) array of int32 with output[j,i,0] = i + origin[0] output[j,i,1] = j + origin[1]

Generates spatial positions for a batch of patches. Args: batch_size: Number of samples in the batch. height: Height