Cerema AWP (Adverse Weather Pedestrian) dataset
收藏资源简介:
Cerema AWP数据集是由法国国家交通、安全与环境工程科学研究中心创建,专注于在恶劣天气条件下对行人进行检测。该数据集包含51,302张图像,所有图像均来自一个完全控制的环境,展示了同一行人在不同天气和光照条件下的情况。数据集的创建旨在评估现有行人检测器在各种条件下的性能,并提供一个用于训练更高效方法的数据集。该数据集适用于多种机器学习任务,如分类、回归和图像生成,特别适用于评估机器学习方法在控制环境下的表现。
The Cerema AWP Dataset was developed by the French National Research Center for Transport, Safety and Environmental Engineering, focusing on pedestrian detection under adverse weather conditions. It contains 51,302 images, all captured in a fully controlled environment, depicting the same pedestrian under varying weather and lighting conditions. The dataset was created to evaluate the performance of existing pedestrian detectors across diverse conditions, as well as to provide a benchmark dataset for training more efficient pedestrian detection methods. It supports a wide range of machine learning tasks, including classification, regression, and image generation, and is particularly well-suited for evaluating the performance of machine learning models in controlled environments.

- 1Baselines and a datasheet for the Cerema AWP dataset法国国家交通、安全与环境工程科学研究中心 · 2018年



