RADIATE (RAdar Dataset In Adverse weaThEr)

RADIATE（不利天气中的雷达数据集）是赫瑞瓦特大学创建的新汽车数据集，包括雷达、激光雷达、立体相机和 GPS/IMU。这些数据是在不同的天气情景（晴天、阴天、夜晚、雾天、雨天和雪天）中收集的，以帮助研究界开发新的车辆感知方法。雷达图像在 7 种不同的场景中进行注释：晴天（停车）、晴天/阴天（城市）、阴天（高速公路）、夜晚（高速公路）、雨天（郊区）、雾（郊区）和雪（郊区）。该数据集包含 8 种不同类型的对象（汽车、货车、卡车、公共汽车、摩托车、自行车、行人和行人组）。

RADIATE (Radar Dataset in Adverse Weather) is a novel automotive dataset created by Heriot-Watt University. It comprises radar, LiDAR, stereo cameras, and GPS/IMU sensors. The dataset was collected across diverse weather conditions including sunny, overcast, nighttime, foggy, rainy, and snowy scenarios, to support the research community in developing innovative vehicle perception methods. Radar imagery within this dataset is annotated across 7 distinct scenarios: sunny (parking lot), sunny/overcast (urban), overcast (highway), nighttime (highway), rainy (suburban), foggy (suburban), and snowy (suburban). Additionally, the dataset covers 8 different object categories: cars, vans, trucks, buses, motorcycles, bicycles, pedestrians, and pedestrian groups.

搜集汇总

背景与挑战

背景概述

RADIATE是一个由赫瑞瓦特大学创建的汽车感知数据集，专注于不利天气条件下的数据收集，包含雷达、激光雷达、立体相机和GPS/IMU等多模态数据，覆盖晴天、阴天、夜晚、雾天、雨天和雪天等多种天气情景。该数据集旨在帮助研究界开发新的车辆感知方法，其雷达图像在7种不同场景中注释，包括8种常见交通对象类型，如汽车、行人和自行车等。