RADIATE: A Radar Dataset for Automotive Perception in Bad Weather

Abstract

Datasets for autonomous cars are essential for the development and benchmarking of perception systems. However, most existing datasets are captured with camera and LiDAR sensors in good weather conditions. In this paper, we present the RAdar Dataset In Adverse weaThEr (RADIATE), aiming to facilitate research on object detection, tracking and scene understanding using radar sensing for safe autonomous driving. RADIATE includes 3 hours of annotated radar images with more than 200K labelled road actors in total, on average about 4.6 instances per radar image. It covers 8 different categories of actors in a variety of weather conditions (e.g., sun, night, rain, fog and snow) and driving scenarios (e.g., parked, urban, motorway and suburban), representing different levels of challenge. To the best of our knowledge, this is the first public radar dataset which provides high-resolution radar images on public roads with a large amount of road actors labelled. The data collected in adverse weather, e.g., fog and snowfall, is unique. Some baseline results of radar based object detection and recognition are given to show that the use of radar data is promising for automotive applications in bad weather, where vision and LiDAR can fail. RADIATE also has stereo images, 32-channel LiDAR and GPS data, directed at other applications such as sensor fusion, localisation and mapping. The public dataset can be accessed at http://pro.hw.ac.uk/radiate/. I. INTRODUCTIONAutonomous driving research and development rely heavily on the use of public datasets in the computer vision and robotics communities [1]-[3]. Camera and LiDAR are the two primary perceptual sensors that are usually adopted. However, since these are visible spectrum sensors, the data is affected dramatically by bad weather, causing attenuation, multiple scattering and

ART-SLAM: Accurate Real-Time 6DoF LiDAR SLAM

“…Lastly, we show a visual evaluation of the accuracy achieved by ART-SLAM on the City 01 Sequence of the RADIATE dataset [16]. This sequence is relatively long and does not contain loop closures, increasing the difficulty of estimating the robot trajectory.…”

“…The proposed system is compared with other methods for point cloud-based SLAM: LOAM [17], LeGO-LOAM [15], A-LOAM, LeGO-LOAM-BOR and HDL [9], with A-LOAM and LeGO-LOAM-BOR being two advanced versions of LeGO-LOAM (code improvement and re-engineering of LeGO-LOAM). We evaluate these systems in four scenarios: three coming from the KITTI dataset [4] [5], corresponding to a short, a medium and a long sequences, respectively, and one from the RADIATE dataset [16], representing a medium sequence with no loops.…”

“…Fig.8. Map corresponding to City Sequence 01 of the RADIATE dataset[16], built by the proposed algorithm.…”

Rethinking of Radar's Role: A Camera-Radar Dataset and Systematic Annotator via Coordinate Alignment

“…However, due to the hardware compatibility and less developed radar perception techniques, most datasets do not incorporate radar signals as a part of their sensor systems. Among the available radar datasets (summarized in Table 1), some of them [8,20,7,29] consider radar data in the format of radar points that do not contain the useful Doppler and surface texture information of objects. Later, researchers start to focus on RF images as the radar data format.…”

Brno urban dataset: Winter extension

“…However, in the future, as autonomous cars will spread worldwide, it will be necessary arXiv:2106.02952v1 [cs.RO] 5 Jun 2021 to cover also these edge conditions. There are several snowcontained recording sessions in the Oxford Dataset [4], in the RADIATE dataset [13], or in the EU Long-Term Dataset [14], but only on a small scale, and to our best knowledge, the only datasets that are focusing on winter conditions are the CADC Dataset by [15], ant the IceVisionSet by [16].…”