Get command-line arguments

Get command-line arguments

Get command-line arguments

gets command line arguments. :return:

Get command-line arguments

Get command-line arguments

Get command-line arguments

Get command-line arguments for object classification training set generation.

Get command-line arguments for KITTI 3D Object ETL

Get command-line arguments

Get command-line arguments

Get command-line arguments

Get command-line arguments

Get command-line arguments

gets command line arguments. :return:

Get centroids Parameters ---------- None Returns ---------- centroids: numpy.ndarray

Get file descriptors for the loggers. They should not be closed when the process is forked.

Get plane RANSAC threshold tau

Get mu

Get object orientation using PCA

Build open3D geometry for PCA Parameters ---------- w: eigenvalues in descending order v: eigenvectors in descending order

Get potential matches Parameters ---------- feature_source: open3d.registration.Feature feature descriptions of source point

Get potential matches Parameters ---------- feature_source: open3d.registration.Feature feature descriptions of source point

Build open3D geometry for surface normals Parameters ---------- normals(numpy.ndarray): surface normals for each point point

Get a function for 3D point -- voxel grid assignment Parameters: points(pandas.DataFrame): points in the point cloud

CONSTANT WEIGHT TODO input: dist (b,n,3) output: weight (b,n,3)

CONSTANT WEIGHT TODO input: dist (b,n,3) output: weight (b,n,3)

Segment ground plane from Velodyne measurement Parameters ---------- data: numpy.ndarray Velodyne measurements as N-by-3 num

Extended request handler This is where WebSocketRequestHandler redirects requests to the new methods. Any sub-classes must call this

Called when a WebSocket ping message is received. This will be called whilst processing recv()/recvmsg(). The default implementation

Called when a WebSocket pong message is received. This will be called whilst processing recv()/recvmsg(). The default implementation

Initial handler for a WebSocket request This method is called when a WebSocket is requested. By default it will create a WebSocket ob

Handle a WebSocket connection. This is called when the WebSocket is ready to be used. A sub-class should perform the necessar

Get label template for extracted classification dataset Returns ---------- labels: dict label container for extracted classi

input: points (b,m,c), idx (b,n,3), weight (b,n,3) output: out (b,n,c)

input: points (b,m,c), idx (b,n,3), weight (b,n,3) output: out (b,n,c)

input: grad_out (b,n,c), idx (b,n,3), weight (b,n,3) output: grad_points (b,m,c)

input: grad_out (b,n,c), idx (b,n,3), weight (b,n,3) output: grad_points (b,m,c)

Check proposal validity using the fast pruning algorithm Parameters ---------- pcd_source: open3d.geometry.PointCloud source

Check proposal validity using the fast pruning algorithm Parameters ---------- pcd_source: open3d.geometry.PointCloud source

Load labels from the extracted object classification dataset Parameters ---------- input_dir: str input path of extracted ob

Load pre-trained object classification network

Run pose estimation on given point cloud pair

Process KITTI 3D Object Parameters ---------- input_dir: str Root path of input KITTI 3D Object dataset output_dir: str

Run point cloud registration on Shenlan dataset

Do something with a WebSockets client connection.

Write a ping message to the WebSocket.

Visualize segmentation results using Open3D Parameters ---------- segmented_cloud: numpy.ndarray Segmented surrounding objec

Run periodically while waiting for connections.

Load trained network and make predictions Parameters ---------- config: dict Model training configuration

Load classification network and predict surrounding object category Parameters ---------- config: dict Model training conf

Load trained network and make predictions Parameters ---------- config: dict Model training configuration

Classify input data Parameters ---------- data: numpy.ndarray Testing set as N-by-D numpy.ndarray

Get cluster labels Parameters ---------- data: numpy.ndarray Testing set as N-by-D numpy.ndarray

Preprocess point cloud with normal: 1. Shift to zero-centered; 2. Downsample / upsample to specified size; 3. (Optional):

Preprocess for classification network