FoundationStereo: Zero-Shot Stereo Matching
Stereo depth model trained on large synthetic data that reports strong zero-shot accuracy on real scenes — a drop-in depth source for stereo-camera robots.在大规模合成数据上训练的双目深度模型，报告称在真实场景上零样本精度很强；可直接作为双目相机机器人的深度来源。
BibTeX
@misc{wen2025foundationstereo,
title = {{FoundationStereo: Zero-Shot Stereo Matching}},
author = {Bowen Wen and Matthew Trepte and Joseph Aribido and Jan Kautz and Orazio Gallo and Stan Birchfield},
year = {2025},
eprint = {2501.09898},
archivePrefix = {arXiv},
primaryClass = {cs.CV},
url = {https://arxiv.org/abs/2501.09898},
}