SAM 3D Artist Objects (SA-3DAO) is a 3D mesh pairing dataset released by Meta Superintelligence Labs in June 2026. The related research paper is... SAM 3D: 3Dfy Anything in ImagesIt aims to evaluate the performance of 3D reconstruction algorithms for the shape and layout of objects in real-world scenarios, and is widely used in performance testing, model optimization, and related research in the field of computer vision for image-to-3D object conversion algorithms. This dataset contains 1,000 pairs of real images paired with 3D meshes handcrafted by professional artists. The dataset covers a wide range of objects, including large buildings, animals, everyday items, and rare objects, and distinguishes between publicly available test samples and competition-specific samples.

Dataset composition:

• Test set: 100 pairs of image/mask-to-3D object pairings for model development and validation.
• Competition set: 900 pairs of 3D object pairings, reserved as benchmarks.

Citation

@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},
}