GOOSE-Ex
收藏资源简介:
GOOSE-Ex数据集是由弗劳恩霍夫光学、系统技术和图像利用研究所创建的,旨在增强自主系统在非结构化户外环境中的感知能力。该数据集包含5000个标注的多模态帧,涵盖了德国各地多种环境条件下的数据。数据集的创建过程包括手动标注和基于平台里程计的帧合并,以提高标注质量。GOOSE-Ex数据集主要应用于自主挖掘机和四足机器人的语义分割任务,旨在解决非结构化环境中感知模型的泛化问题。
The GOOSE-Ex dataset was developed by the Fraunhofer Institute for Optronics, System Technologies and Image Exploitation to enhance the perception capabilities of autonomous systems in unstructured outdoor environments. It contains 5,000 annotated multimodal frames, encompassing data collected under various environmental conditions across Germany. The dataset construction process includes manual annotation and frame merging based on platform odometry to improve annotation quality. The GOOSE-Ex dataset is primarily utilized for semantic segmentation tasks targeting autonomous excavators and quadruped robots, with the core objective of addressing the generalization challenge of perception models in unstructured environments.

- 1Excavating in the Wild: The GOOSE-Ex Dataset for Semantic Segmentation弗劳恩霍夫光学、系统技术和图像利用研究所 · 2024年



