nuImages
收藏OpenDataLab2026-05-17 更新2024-05-09 收录
下载链接:
https://opendatalab.org.cn/OpenDataLab/nuImages
下载链接
链接失效反馈官方服务:
资源简介:
用于自动驾驶的 nuScenes 数据集的发布取得了压倒性的成功。该数据集被 8,000 多名用户下载,论文在最初发布后的 15 个月内被引用 220 次。在 nuScenes 发布之后,出现了大量类似的数据集发布,例如来自 Lyft、Waymo、Hesai、Argo、Audi、Ford 和 Zoox。
nuScenes 的优势在于 1000 个精心策划的带有 3d 注释的场景,涵盖了许多具有挑战性的驾驶情况。 nuImages 通过提供来自更大数据池的 93,000 个 2d 注释图像来补充此产品。我们还提供过去和未来的相机图像,总共有 1,200,000 张相机图像。
The release of the nuScenes dataset for autonomous driving has witnessed overwhelming success. It has been downloaded by over 8,000 users, and its associated research paper has been cited more than 220 times within 15 months following its initial release. Shortly after the public launch of nuScenes, a plethora of analogous autonomous driving datasets have been published, including those from Lyft, Waymo, Hesai, Argo, Audi, Ford, and Zoox.
The core advantage of nuScenes lies in its 1,000 carefully curated scenes with 3D annotations, covering a wide range of challenging driving scenarios. The nuImages dataset complements this resource by offering 93,000 2D-annotated images sourced from a larger data corpus. Additionally, we provide both past and future camera imagery, amounting to a total of 1,200,000 camera images.
提供机构:
OpenDataLab
创建时间:
2022-03-17
搜集汇总
数据集介绍

背景与挑战
背景概述
nuImages是nuScenes数据集的补充,专注于自动驾驶领域,提供93,000张带有2D注释的图像和总计1,200,000张相机图像,涵盖过去和未来的场景。该数据集包含语义分割和物体检测任务,使用CC BY-NC-SA 4.0许可证,由APTIV Company于2020年发布,旨在支持自动驾驶研究和开发。
以上内容由遇见数据集搜集并总结生成



