GFrames: Gradient-Based Local Reference Frame for 3D Shape Matching
Abstract
We introduce GFrames, a novel local reference frame (LRF) construction for 3D meshes and point clouds. GFrames are based on the computation of the intrinsic gradient of a scalar field defined on top of the input shape. The resulting tangent vector field defines a repeatable tangent direction of the local frame at each point; importantly, it directly inherits the properties and invariance classes of the underlying scalar function, making it remarkably robust under strong sampling artifacts, vertex noise, as well as non-rigid deformations. Existing local descriptors can directly benefit from our repeatable frames, as we showcase in a selection of 3D vision and shape analysis applications where we demonstrate state-of-the-art performance in a variety of challenging settings.
Cite
Text
Melzi et al. "GFrames: Gradient-Based Local Reference Frame for 3D Shape Matching." Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2019. doi:10.1109/CVPR.2019.00476Markdown
[Melzi et al. "GFrames: Gradient-Based Local Reference Frame for 3D Shape Matching." Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2019.](https://mlanthology.org/cvpr/2019/melzi2019cvpr-gframes/) doi:10.1109/CVPR.2019.00476BibTeX
@inproceedings{melzi2019cvpr-gframes,
title = {{GFrames: Gradient-Based Local Reference Frame for 3D Shape Matching}},
author = {Melzi, Simone and Spezialetti, Riccardo and Tombari, Federico and Bronstein, Michael M. and Di Stefano, Luigi and Rodola, Emanuele},
booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition},
year = {2019},
doi = {10.1109/CVPR.2019.00476},
url = {https://mlanthology.org/cvpr/2019/melzi2019cvpr-gframes/}
}