Write a pong message to the WebSocket.

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

Process one sample from KITTI 3D Object Parameters ---------- index: int KITTI 3D Object measurement ID input_dir: str

:param path: :return: [x,y,z,nx,ny,nz]: 6xN

:param path: :return: homography matrix of the point cloud, N*3

resolve host with also IP address parsing */

Radius based nearest neighbor search in a faster way Parameters ---------- query : numpy.ndarray query p

Radius based nearest neighbor search Parameters ---------- query : numpy.ndarray query point rnn

Radius based nearest neighbor search Parameters ---------- query : numpy.ndarray query point rnn

Solve ICP Parameters ---------- P: numpy.ndarray source point cloud as N-by-3 numpy.ndarray Q: numpy.ndarray tar

Solve ICP Parameters ---------- P: numpy.ndarray source point cloud as N-by-3 numpy.ndarray Q: numpy.ndarray tar

Serialize Parameters ---------- output_name: str Output TFRecord name

Build KDTree

Build KDTree

Whether the query contains the given octant Parameters ---------- query: numpy.ndarray query point

Node for KDTree Parameters ---------- depth : int Current depth point_indices: list of int

Node for KDTree Parameters ---------- axis : int Splitting axis depth : int Current

Get relative path of example Parameters ---------- filename_example: str Short filename of example.

Get relative path of example Parameters ---------- filename_example: str Short filename of example.

Get initial centroids using KMeans++ selection Parameters ---------- data: numpy.ndarray Training set as

Generate next axis for KDTree splitting using round-robin Parameters ---------- axis: int splitting ind

Set initial GMM params with K-Means initialization Parameters ---------- data: numpy.ndarray Training se

Whether the query is inside given octant Parameters ---------- query: numpy.ndarray query point

Load labels Parameters ---------- filename_labels: str Filename of dataset labels.

Load labels Parameters ---------- filename_labels: str Filename of dataset labels.

Whether the query overlaps with given octant Parameters ---------- query: numpy.ndarray query point

Sort key by value Parameters ---------- key: numpy.ndarray value indices value: numpy.ndarray

Return a tolerance which is independent of the dataset

assign given 3D point to voxel grid Parameters: x(float): X y(float): Y z(float): Z Ret

Segment surrounding objects using DBSCAN Parameters ---------- data: numpy.ndarray Segmented point cloud as N-by-3 numpy.nda

Colormap for segmentation result Parameters ---------- c: int Cluster ID C

Describe the selected keypoint using Fast Point Feature Histogram(FPFH) Parameters ---------- point_cloud: Open3D.geometry.PointClou

Detect point cloud key points using Intrinsic Shape Signature(ISS) Parameters ---------- point_cloud: Open3D.geometry.PointCloud

Proxy client WebSocket to normal target socket.

Visualize the relationship between measurement counts and object-ego vehicle distance Parameters ---------- labels: dict of pandas.D

Filter point measurements using bounding box in object frame Parameters ---------- X: numpy.ndarray Points in object frame

Estimate the K centroids Parameters ---------- data: numpy.ndarray Training set as N-by-D numpy.ndarray

Generate dataset for spectral clustering Parameters ---------- visualize: boolean Whether to visualize the generated data

Generate train-test split Parameters ---------- input_dir: str Root path of resampled training set Returns --------

Generate training set Parameters ---------- labels: dict of pandas.DataFrame labels of the extracted dataset input_dir: