SAM 3D Artist Objects (SA-3DAO) Dataset

Overview

The SAM 3D Artist Objects (SA-3DAO) dataset is an evaluation benchmark of 1,000 image and 3D mesh pairs designed for measuring progress in real-world 3D reconstruction of object shape and layout. The objects span a wide range of categories – from large structures like churches and ski lifts to animals, everyday household items, and rare objects – each paired with the real-world image in which it naturally appears. Professional 3D artists created image-aligned 3D meshes by hand, representing an expert human upper bound for visually grounded 3D reconstruction. The dataset was introduced in our paper "SAM 3D: 3Dfy Anything in Images".

We release 100 objects with image/mask and 3D object pairs for testing and development, and withhold 900 3D objects for benchmark competition. Compete with your model in the leaderboard on HuggingFace!

License

CC by NC licenses (https://creativecommons.org/licenses/by-nc/4.0/deed.en)

Citation

If you use this dataset, or the accompanying processed data, models, or code, please cite the following BibTeX entry:

@article{sam3d3dfyimages,
 title={SAM 3D: 3Dfy Anything in Images},
 author={SAM 3D Team and Xingyu Chen
 and Fu-Jen Chu and Pierre Gleize
 and Kevin J Liang and Alexander Sax
 and Hao Tang and Weiyao Wang
 and Michelle Guo and Thibaut Hardin
 and Xiang Li and Aohan Lin and Jiawei Liu
 and Ziqi Ma and Anushka Sagar and Bowen Song
 and Xiaodong Wang and Jianing Yang and Bowen Zhang
 and Piotr Dollár and Georgia Gkioxari
 and Matt Feiszli and Jitendra Malik},
 year={2025},
 eprint={2511.16624},
 archivePrefix={arXiv},
 primaryClass={cs.CV},
 url={https://arxiv.org/abs/2511.16624},
 }