WAFT-Stereo: Warping-Alone Field Transforms for Stereo Matching

WAFT-Stereo is a simple and effective warping-based method for stereo matching. It demonstrates that cost volumes, a common design used in many leading methods, are not necessary for strong performance and can be replaced by warping with improved efficiency.

WAFT-Stereo ranks first on ETH3D, KITTI and Middlebury public benchmarks, reducing the zero-shot error by 81% on ETH3D benchmark, while being 1.8-6.7x faster than competitive methods.

Citation

If you find WAFT-Stereo useful for your work, please consider citing the paper:

@misc{wang2026waftstereowarpingalonefieldtransforms,
      title={WAFT-Stereo: Warping-Alone Field Transforms for Stereo Matching}, 
      author={Yihan Wang and Jia Deng},
      year={2026},
      eprint={2603.24836},
      archivePrefix={arXiv},
      primaryClass={cs.CV},
      url={https://arxiv.org/abs/2603.24836}, 
}
Downloads last month

-

Downloads are not tracked for this model. How to track
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Paper for MemorySlices/WAFT-Stereo