RADDet (Range-Azimuth-Doppler based Radar Dataset)

RADDet 是一个雷达数据集，其中包含 Range-Azimuth-Doppler 张量形式的雷达数据，以及用于动态道路用户的张量边界框、类别标签和笛卡尔鸟瞰图范围图上的 2D 边界框。它用于训练和评估使用汽车雷达进行目标检测的方法。RADDet is a radar dataset containing radar data in the form of Range-Azimuth-Doppler tensors, as well as tensor-based bounding boxes, category labels, and 2D bounding boxes on Cartesian bird's-eye view range maps for dynamic road users. It is used for training and evaluating object detection methods that leverage automotive radars.

搜集汇总

背景与挑战

背景概述

RADDet是一个基于Range-Azimuth-Doppler张量的雷达数据集，专门用于动态道路用户的目标检测，包含雷达数据、边界框和类别标签。该数据集旨在支持汽车雷达目标检测方法的训练和评估，由渥太华大学和Sensor Cortek于2021年发布。