Turn the depth map into a 3D point cloud in world space Args: depth: (b, v, h, w, 1) intrinsics: (b, v, 3, 3) c2w: (

Handle file uploads and update gallery. Args: input_video: Path to input video file input_images: List of in

Update two config dictionaries recursively. dict1 get masked by dict2, and we retuen dict1. Args: dict1 (dict): first dictionary to

Update view selector dropdowns based on available views. Args: processed_data: Processed data dictionary Return

Color-map the depth map.

Visualize a depth map using a colormap. Args: depth: Input depth map array depth_min: Minimum depth value for normalization.

point_map2 -> point_map1

see: src/colmap/scene/reconstruction.cc void Reconstruction::WriteCamerasBinary(const std::string& path) void Reconstruction::Rea

see: src/colmap/scene/reconstruction.cc void Reconstruction::WriteCamerasText(const std::string& path) void Reconstruction::ReadC

see: src/colmap/scene/reconstruction.cc void Reconstruction::ReadImagesBinary(const std::string& path) void Reconstruction::Write

see: src/colmap/scene/reconstruction.cc void Reconstruction::ReadImagesText(const std::string& path) void Reconstruction::WriteIm

see: src/colmap/scene/reconstruction.cc void Reconstruction::ReadPoints3DBinary(const std::string& path) void Reconstruction::Wri

see: src/colmap/scene/reconstruction.cc void Reconstruction::ReadPoints3DText(const std::string& path) void Reconstruction::Write

Convert SE3 to Sim3 (add unit scale)

Convert model output to Prediction object. Args: model_output: Model output dictionary containing depth, conf, extrinsic

Returns: (tensor, extrinsics_list, intrinsics_list) tensor shape: (1, N, 3, H, W)

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

Lazy import to avoid circular import when running as __main__.

Initialize the loop detector Args: image_dir: Directory path containing images ckpt_path: Model checkpoint path

Initialize DepthAnything3 with specified preset. Args: model_name: The name of the model preset to use.

Initialize the output processor.

Initialize NestedDepthAnything3Net with two branches. Args: preset: Configuration for main depth estimation branch

Args: img_size (int, tuple): input image size patch_size (int, tuple): patch size in_chans (int): number

Initializes the 2D RoPE module.

Initialize the application. Args: model_dir: Path to the model directory workspace_dir: Path to the workspac

Initialize the UI components handler.

Initialize the event handlers.

Initialize the model inference handler.

Initialize the file handler.

Initialize the visualization handler.

Initialize the evaluator. Args: work_dir: Base directory for model outputs and metric files datas: List of d

Initialize the printer. Args: use_color: Whether to use ANSI colors in output

Denormalize to [0,255] and output (N, H, W, 3) uint8.

Directory for storing metric JSON files.

Run inference task in background thread with OOM protection.

point_map2 -> point_map1

Apply Sim(3) (R, t, s) to a batch of world-to-camera extrinsics ext_est. Returns the aligned extrinsics, with the same shape as input.

Automatically detect input type and run appropriate processing. Supports: - Single image file (.jpg, .png, etc.) - Directory of imag

Start model backend service with integrated gallery.

Convert camera space points to pixel space points. Args: camera_space_points (torch.Tensor): Camera space points with shape (b, v1,