- P
ScanNet Dataset
- paperswithcode.com
- opendatalab.com
Updated Nov 25, 2019
Share
Facebook
Twitter
EmailClick to copy linkLink copiedCiteAngela Dai; Angel X. Chang; Manolis Savva; Maciej Halber; Thomas Funkhouser; Matthias Nießner (2019). ScanNet Dataset [Dataset]. https://paperswithcode.com/dataset/scannetDataset updatedNov 25, 2019AuthorsAngela Dai; Angel X. Chang; Manolis Savva; Maciej Halber; Thomas Funkhouser; Matthias NießnerDescriptionScanNet is an instance-level indoor RGB-D dataset that includes both 2D and 3D data. It is a collection of labeled voxels rather than points or objects. Up to now, ScanNet v2, the newest version of ScanNet, has collected 1513 annotated scans with an approximate 90% surface coverage. In the semantic segmentation task, this dataset is marked in 20 classes of annotated 3D voxelized objects.
- h
scannet
- huggingface.co
Updated Apr 18, 2024+ more versionsShareFacebookTwitterEmailClick to copy linkLink copiedCiteyuchen zhou (2024). scannet [Dataset]. https://huggingface.co/datasets/yuchen0187/scannetCroissantCroissant is a format for machine-learning datasets. Learn more about this at mlcommons.org/croissant.Dataset updatedApr 18, 2024Authorsyuchen zhouDescriptionyuchen0187/scannet dataset hosted on Hugging Face and contributed by the HF Datasets community
- f
Quantitative result with multi-view RGB inputs is evaluated on the val-set...
- plos.figshare.com
xlsUpdated May 19, 2025ShareFacebookTwitterEmailClick to copy linkLink copiedCiteGang Xu; Biao Leng; Zhang Xiong (2025). Quantitative result with multi-view RGB inputs is evaluated on the val-set of ScanNet-V2. [Dataset]. http://doi.org/10.1371/journal.pone.0320589.t001xlsAvailable download formatsUnique identifierhttps://doi.org/10.1371/journal.pone.0320589.t001Dataset updatedMay 19, 2025Dataset provided byPLOS ONEAuthorsGang Xu; Biao Leng; Zhang XiongLicenseAttribution 4.0 (CC BY 4.0)https://creativecommons.org/licenses/by/4.0/
License information was derived automaticallyDescriptionQuantitative result with multi-view RGB inputs is evaluated on the val-set of ScanNet-V2.
- h
Multi3DRefer
- huggingface.co
Updated Jun 26, 2025ShareFacebookTwitterEmailClick to copy linkLink copiedCite3D Language & Generation Research Group (2025). Multi3DRefer [Dataset]. https://huggingface.co/datasets/3dlg-hcvc/Multi3DReferDataset updatedJun 26, 2025Dataset authored and provided by3D Language & Generation Research GroupLicenseMIT Licensehttps://opensource.org/licenses/MIT
License information was derived automaticallyDescriptionMulti3DRefer Dataset
This repository contains the Multi3DRefer dataset, introduced in Multi3DRefer: Grounding Text Description to Multiple 3D Objects.
Dataset Attributesscene_id: (str) the scene identifier as defined in ScanNetv2. object_name: (str) the name of the target object(s). ann_id: (int) the local annotation ID within a scene. description: (str) the language description of the target object(s) in the scene. object_ids: (List[int]) the target object ID(s) as… See the full description on the dataset page: https://huggingface.co/datasets/3dlg-hcvc/Multi3DRefer.
Not seeing a result you expected?
Learn how you can add new datasets to our index.
FacebookTwitterEmailScanNet is an instance-level indoor RGB-D dataset that includes both 2D and 3D data. It is a collection of labeled voxels rather than points or objects. Up to now, ScanNet v2, the newest version of ScanNet, has collected 1513 annotated scans with an approximate 90% surface coverage. In the semantic segmentation task, this dataset is marked in 20 classes of annotated 3D voxelized objects.