Neural-Pull: Learning Signed Distance Function from Point clouds by Learning to Pull Space onto Surface

Baorui Ma, Zhizhong Han, Yu-Shen Liu, Matthias Zwicker
Proceedings of the 38th International Conference on Machine Learning, PMLR 139:7246-7257, 2021.

Abstract

Reconstructing continuous surfaces from 3D point clouds is a fundamental operation in 3D geometry processing. Several recent state-of-the-art methods address this problem using neural networks to learn signed distance functions (SDFs). In this paper, we introduce Neural-Pull, a new approach that is simple and leads to high quality SDFs. Specifically, we train a neural network to pull query 3D locations to their closest points on the surface using the predicted signed distance values and the gradient at the query locations, both of which are computed by the network itself. The pulling operation moves each query location with a stride given by the distance predicted by the network. Based on the sign of the distance, this may move the query location along or against the direction of the gradient of the SDF. This is a differentiable operation that allows us to update the signed distance value and the gradient simultaneously during training. Our outperforming results under widely used benchmarks demonstrate that we can learn SDFs more accurately and flexibly for surface reconstruction and single image reconstruction than the state-of-the-art methods. Our code and data are available at https://github.com/mabaorui/NeuralPull.

Cite this Paper


BibTeX
@InProceedings{pmlr-v139-ma21b, title = {Neural-Pull: Learning Signed Distance Function from Point clouds by Learning to Pull Space onto Surface}, author = {Ma, Baorui and Han, Zhizhong and Liu, Yu-Shen and Zwicker, Matthias}, booktitle = {Proceedings of the 38th International Conference on Machine Learning}, pages = {7246--7257}, year = {2021}, editor = {Meila, Marina and Zhang, Tong}, volume = {139}, series = {Proceedings of Machine Learning Research}, month = {18--24 Jul}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v139/ma21b/ma21b.pdf}, url = {https://proceedings.mlr.press/v139/ma21b.html}, abstract = {Reconstructing continuous surfaces from 3D point clouds is a fundamental operation in 3D geometry processing. Several recent state-of-the-art methods address this problem using neural networks to learn signed distance functions (SDFs). In this paper, we introduce Neural-Pull, a new approach that is simple and leads to high quality SDFs. Specifically, we train a neural network to pull query 3D locations to their closest points on the surface using the predicted signed distance values and the gradient at the query locations, both of which are computed by the network itself. The pulling operation moves each query location with a stride given by the distance predicted by the network. Based on the sign of the distance, this may move the query location along or against the direction of the gradient of the SDF. This is a differentiable operation that allows us to update the signed distance value and the gradient simultaneously during training. Our outperforming results under widely used benchmarks demonstrate that we can learn SDFs more accurately and flexibly for surface reconstruction and single image reconstruction than the state-of-the-art methods. Our code and data are available at https://github.com/mabaorui/NeuralPull.} }
Endnote
%0 Conference Paper %T Neural-Pull: Learning Signed Distance Function from Point clouds by Learning to Pull Space onto Surface %A Baorui Ma %A Zhizhong Han %A Yu-Shen Liu %A Matthias Zwicker %B Proceedings of the 38th International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2021 %E Marina Meila %E Tong Zhang %F pmlr-v139-ma21b %I PMLR %P 7246--7257 %U https://proceedings.mlr.press/v139/ma21b.html %V 139 %X Reconstructing continuous surfaces from 3D point clouds is a fundamental operation in 3D geometry processing. Several recent state-of-the-art methods address this problem using neural networks to learn signed distance functions (SDFs). In this paper, we introduce Neural-Pull, a new approach that is simple and leads to high quality SDFs. Specifically, we train a neural network to pull query 3D locations to their closest points on the surface using the predicted signed distance values and the gradient at the query locations, both of which are computed by the network itself. The pulling operation moves each query location with a stride given by the distance predicted by the network. Based on the sign of the distance, this may move the query location along or against the direction of the gradient of the SDF. This is a differentiable operation that allows us to update the signed distance value and the gradient simultaneously during training. Our outperforming results under widely used benchmarks demonstrate that we can learn SDFs more accurately and flexibly for surface reconstruction and single image reconstruction than the state-of-the-art methods. Our code and data are available at https://github.com/mabaorui/NeuralPull.
APA
Ma, B., Han, Z., Liu, Y. & Zwicker, M.. (2021). Neural-Pull: Learning Signed Distance Function from Point clouds by Learning to Pull Space onto Surface. Proceedings of the 38th International Conference on Machine Learning, in Proceedings of Machine Learning Research 139:7246-7257 Available from https://proceedings.mlr.press/v139/ma21b.html.

Related Material