Datasetpublic

manycore-research/SpatialLM-Dataset

SpatialLM Dataset The SpatialLM dataset is a large-scale, high-quality synthetic dataset designed by professional 3D designers and used for real-world production. It contains point clouds from 12,328 diverse indoor scenes comprising 54,778 rooms, each paired with rich ground-truth 3D annotations. SpatialLM dataset provides an additional valuable resource for advancing research in indoor scene understanding, 3D perception… See the full description on the dataset page: https://huggingface.co/datasets/manycore-research/SpatialLM-Dataset.

sourceHugging Facecc-by-nc-4.0updated 1y agoView on Hugging Face

15likes1.4kdownloads

rootdataset_info.json

dataset_info.json26 linesDownload Raw Back to root

1{2  "spatiallm_train": {3    "file_name": "spatiallm_train.json",4    "formatting": "sharegpt",5    "columns": {6      "messages": "conversations",7      "point_clouds": "point_clouds"8    }9  },10  "spatiallm_val": {11    "file_name": "spatiallm_val.json",12    "formatting": "sharegpt",13    "columns": {14      "messages": "conversations",15      "point_clouds": "point_clouds"16    }17  },18  "spatiallm_test": {19    "file_name": "spatiallm_test.json",20    "formatting": "sharegpt",21    "columns": {22      "messages": "conversations",23      "point_clouds": "point_clouds"24    }25  }26}