mateoguaman/vamos_dataset
收藏Hugging Face2025-10-23 更新2025-10-25 收录
下载链接：
https://hf-mirror.com/datasets/mateoguaman/vamos_dataset
下载链接
链接失效反馈官方服务：
资源简介：
VAMOS导航数据集（包含视觉语言联合训练数据）结合了多个公开可用的数据集和作者收集的Spot数据。这个版本包含了导航数据以及来自COCO-QA和Localized Narratives的联合训练数据。数据集包括TartanDrive 2数据的100%，SCAND数据的50%，CODa数据的25%，作者收集的Spot数据的100%，以及用于辅助VQA和语言定位任务的COCO-QA和Localized Narratives。
The VAMOS Navigation Dataset (with visual-language co-training data) combines multiple publicly available datasets and in-domain Spot data collected by the authors. This version includes navigation data as well as co-training data from COCO-QA and Localized Narratives. The dataset includes 100% of TartanDrive 2 data, 50% of SCAND data, 25% of CODa data, 100% of in-domain Spot data (collected by the authors), and COCO-QA and Localized Narratives (used for auxiliary VQA and language grounding tasks).
提供机构：
mateoguaman
