Radar Artifact Labeling Framework (RALF)
收藏arXiv2020-12-03 更新2024-08-06 收录
下载链接：
http://arxiv.org/abs/2012.01993v1
下载链接
链接失效反馈官方服务：
资源简介：
Radar Artifact Labeling Framework (RALF) 是由德国的保时捷股份公司和卡尔斯鲁厄理工学院共同开发的数据集，专注于自动驾驶中的雷达数据处理。该数据集包含328万个点，用于自动标记汽车雷达数据，区分真实目标与伪影。RALF利用车载传感器，通过光学感知和时间信号分析两个并行处理流程，自动生成雷达原始检测的合理性标签。此数据集主要应用于自动驾驶和高级驾驶辅助系统的感知学习任务，旨在解决雷达数据处理中的噪声和伪影问题。
Radar Artifact Labeling Framework (RALF) is a dataset co-developed by Porsche AG of Germany and Karlsruhe Institute of Technology, focusing on radar data processing in autonomous driving. This dataset contains 3.28 million data points, designed for automated labeling of automotive radar data to distinguish between real targets and artifacts. RALF utilizes on-board sensors and automatically generates validity labels for raw radar detections via two parallel processing pipelines: optical perception and temporal signal analysis. This dataset is mainly applied to perception learning tasks for autonomous driving and Advanced Driver Assistance Systems (ADAS), aiming to solve the noise and artifact issues in radar data processing.
提供机构：
保时捷股份公司
创建时间：
2020-12-03
