Input: pc: numpy array (N,C), first 3 channels are XYZ z is facing forward, x is left ward, y is downward rot_angle:

Takes an object and a projection matrix (P) and projects the 3d bounding box into the image plane. Returns: corners_3d: (

cx, cy, cz, l, w, h, ry

Filter lidar points, keep those in image FOV

Input: nx3 first two channels are uv, 3rd channel is depth in rect camera coord. Output: nx3 points in rect camera coord.

pc: (N,3), box3d: (8,3)

pc: (N,3), box3d: (8,3)

b, 7 (cx, cy, cz, l, w, h, r)

Frustum rotation of point clouds. NxC points with first 3 channels as XYZ z is facing forward, x is left ward, y is downward

Get the frustum rotation angle, it isshifted by pi/2 so that it can be directly used to adjust GT heading angle

Get the frustum rotation angle, it isshifted by pi/2 so that it can be directly used to adjust GT heading angle

Frustum rotation of point clouds. NxC points with first 3 channels as XYZ z is facing forward, x is left ward, y is downward

method: xavier_uniform, kaiming_normal, constant

Ref: http://stackoverflow.com/questions/24467972/calculate-area-of-polygon-given-x-y-coordinates

Input: nx3 first two channels are uv, 3rd channel is depth in rect camera coord. Output: nx3 points in rect camera coo

Input: (N,3) Output: (N,2) UV and (N,) depth

Call this function in your script after you have finished setting all cfg values that are necessary (e.g., merging a config from a file, merging

anchors: (N, 4) ndarray of float gt_boxes: (K, 4) ndarray of float overlaps: (N, K) ndarray of overlap between boxes and query_boxes

Input: nx3 points in Cartesian Oupput: nx4 points in Homogeneous by pending 1

Compute area of two convex hull's intersection area. p1,p2 are a list of (x,y) tuples of hull vertices. return a list of (x,y) for th

8, 3

Load a yaml config file and merge it into the global config.

Merge config keys, values in a list (e.g., from command line) into the global config. For example, `cfg_list = ['TEST.NMS', 0.5]`.

Input: nx3 points in rect camera coord. Output: nx3 points in velodyne coord.

boxes_2d: n, 4 (cx, cz, l, w, r) return n, 4, 2

Read in a calibration file and parse into a dictionary. Ref: https://github.com/utiasSTARS/pykitti/blob/master/pykitti/utils.py

Parse lines in 2D detection output files

Checks that `value_a`, which is intended to replace `value_b` is of the right type. The type is correct if it matches exactly or is one of a few

Decodes a raw config value (e.g., from a yaml config files or command line argument) into a Python object.

Merge config dictionary a into config dictionary b, clobbering the options in b whenever they are also specified in a.

Compute 3D bounding box IoU. Input: corners1: numpy array (8,3), assume up direction is negative Y corners2: numpy array (8,3),

corners: (8,3) no assumption on axis direction

b, 7 (cx, cy, cz, l, w, h, r)

Extract point clouds and corresponding annotations in frustums defined generated from 2D bounding boxes Lidar points and 3d boxes are

Extract point clouds and corresponding annotations in frustums defined generated from 2D bounding boxes Lidar points and 3d boxes are

Extract point clouds in frustums extruded from 2D detection boxes. Update: Lidar points and 3d boxes are in *rect camera* coord system

pc: (N,3), box3d: (8,3)

Flip X-right,Y-forward,Z-up to X-right,Y-down,Z-forward Input and output are both (N,3) array

Return a boolean indicating whether a flag name is supported on the specified compiler.

Clip a polygon with another polygon. Ref: https://rosettacode.org/wiki/Sutherland-Hodgman_polygon_clipping#Python Args: subjectPolygo

Project 3d points to image plane. Usage: pts_2d = projectToImage(pts_3d, P) input: pts_3d: nx3 matrix P: 3x4 projection

Randomly shift box center, randomly scale width and height

Randomly shift box center, randomly scale width and height

boxes_2d: n, 4 (cx, cz, l, w, r) return n, 4, 2

boxes_3d, qboxes_3d: (cx, cy, cz, l, w, h, r) n, 7