Open X-Embodiment
1M+ real-robot trajectories from 20+ institutions.
GRAID NuImages Question-Answer Dataset Paper Project Page Authors: Charles Xu, Qiyang Li, Jianlan Luo, Sergey Levine Overview This dataset was generated using GRAID (Generating Reasoning questions from Analysis of Images via Discriminative artificial intelligence), a framework for creating spatial reasoning datasets from object detection annotations. GRAID transforms raw object detection data into structured question-answer pairs that test various aspects of object… See the full description on the dataset page:
Source: Hugging Face Hub (kd7/graid-nuimages). Metadata imported from the dataset’s Hub tags.