This class represents a set of bounding boxes. The bounding boxes are represented as a Nx4 Tensor. In order to uniquely determine the bou

holding bounding boxes for ground truth and detections

holding data needed for precision-recall and precision-aos

Load and parse object data into a usable format.

Lighting noise(AlexNet - style PCA - based noise)

Lighting noise(AlexNet - style PCA - based noise)

3d object label

evaluated object classes

Calibration matrices and utils 3d XYZ in <label>.txt are in rect camera coord. 2d box xy are in image2 coord Points in <lidar

Calibration matrices and utils 3d XYZ in <label>.txt are in rect camera coord. 2d box xy are in image2 coord Points in <lidar

easy, moderate and hard evaluation level

BatchNorm2d where the batch statistics and the affine parameters are fixed

evaluation metrics: image, ground or 3D

Composes several transforms together in random order.

Composes several transforms together in random order.

holding detection data

holding ground truth data