makeabilitylab/disabilityparking
收藏Hugging Face2025-09-20 更新2025-11-01 收录
下载链接：
https://hf-mirror.com/datasets/makeabilitylab/disabilityparking
下载链接
链接失效反馈官方服务：
资源简介：
AccessParkCV是一个基于深度学习的管道，用于从正射校正的航空影像中检测和表征残疾人停车位的宽度。我们发布了一个包含7,069个标记的停车位（和4,693个标记的通行道）的数据集，用于训练使AccessParkCV成为可能的模型。该数据集具有8个类别，包括objects、access_aisle、curbside、dp_no_aisle、dp_one_aisle、dp_two_aisle、one_aisle和two_aisle。数据集包含训练、测试和验证三个划分，分别包含3,688、717和720个样本。
AccessParkCV is a deep learning pipeline that detects and characterizes the width of disability parking spaces from orthorectified aerial imagery. We publish a dataset of 7,069 labeled parking spaces (and 4,693 labeled access aisles), which we used to train the models making AccessParkCV possible. The dataset has 8 classes including objects, access_aisle, curbside, dp_no_aisle, dp_one_aisle, dp_two_aisle, one_aisle, and two_aisle. The dataset is split into train, test, and validation sets with 3,688, 717, and 720 examples respectively.
提供机构：
makeabilitylab
