NeuralIndicator: Implicit Surface Reconstruction from Neural Indicator Priors

Shi-Sheng Huang, Guo Chen, Chen Li Heng, Hua Huang
Proceedings of the 41st International Conference on Machine Learning, PMLR 235:19622-19634, 2024.

Abstract

The neural implicit surface reconstruction from unorganized points is still challenging, especially when the point clouds are incomplete and/or noisy with complex topology structure. Unlike previous approaches performing neural implicit surface learning relying on local shape priors, this paper proposes to utilize global shape priors to regularize the neural implicit function learning for more reliable surface reconstruction. To this end, we first introduce a differentiable module to generate a smooth indicator function, which globally encodes both the indicative prior and local SDFs of the entire input point cloud. Benefit from this, we propose a new framework, called NeuralIndicator, to jointly learn both the smooth indicator function and neural implicit function simultaneously, using the global shape prior encoded by smooth indicator function to effectively regularize the neural implicit function learning, towards reliable and high-fidelity surface reconstruction from unorganized points without any normal information. Extensive evaluations on synthetic and real-scan datasets show that our approach consistently outperforms previous approaches, especially when point clouds are incomplete and/or noisy with complex topology structure.

Cite this Paper

BibTeX
@InProceedings{pmlr-v235-huang24b, title = {{N}eural{I}ndicator: Implicit Surface Reconstruction from Neural Indicator Priors}, author = {Huang, Shi-Sheng and Chen, Guo and Heng, Chen Li and Huang, Hua}, booktitle = {Proceedings of the 41st International Conference on Machine Learning}, pages = {19622--19634}, year = {2024}, editor = {Salakhutdinov, Ruslan and Kolter, Zico and Heller, Katherine and Weller, Adrian and Oliver, Nuria and Scarlett, Jonathan and Berkenkamp, Felix}, volume = {235}, series = {Proceedings of Machine Learning Research}, month = {21--27 Jul}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v235/main/assets/huang24b/huang24b.pdf}, url = {https://proceedings.mlr.press/v235/huang24b.html}, abstract = {The neural implicit surface reconstruction from unorganized points is still challenging, especially when the point clouds are incomplete and/or noisy with complex topology structure. Unlike previous approaches performing neural implicit surface learning relying on local shape priors, this paper proposes to utilize global shape priors to regularize the neural implicit function learning for more reliable surface reconstruction. To this end, we first introduce a differentiable module to generate a smooth indicator function, which globally encodes both the indicative prior and local SDFs of the entire input point cloud. Benefit from this, we propose a new framework, called NeuralIndicator, to jointly learn both the smooth indicator function and neural implicit function simultaneously, using the global shape prior encoded by smooth indicator function to effectively regularize the neural implicit function learning, towards reliable and high-fidelity surface reconstruction from unorganized points without any normal information. Extensive evaluations on synthetic and real-scan datasets show that our approach consistently outperforms previous approaches, especially when point clouds are incomplete and/or noisy with complex topology structure.} }
Endnote
%0 Conference Paper %T NeuralIndicator: Implicit Surface Reconstruction from Neural Indicator Priors %A Shi-Sheng Huang %A Guo Chen %A Chen Li Heng %A Hua Huang %B Proceedings of the 41st International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2024 %E Ruslan Salakhutdinov %E Zico Kolter %E Katherine Heller %E Adrian Weller %E Nuria Oliver %E Jonathan Scarlett %E Felix Berkenkamp %F pmlr-v235-huang24b %I PMLR %P 19622--19634 %U https://proceedings.mlr.press/v235/huang24b.html %V 235 %X The neural implicit surface reconstruction from unorganized points is still challenging, especially when the point clouds are incomplete and/or noisy with complex topology structure. Unlike previous approaches performing neural implicit surface learning relying on local shape priors, this paper proposes to utilize global shape priors to regularize the neural implicit function learning for more reliable surface reconstruction. To this end, we first introduce a differentiable module to generate a smooth indicator function, which globally encodes both the indicative prior and local SDFs of the entire input point cloud. Benefit from this, we propose a new framework, called NeuralIndicator, to jointly learn both the smooth indicator function and neural implicit function simultaneously, using the global shape prior encoded by smooth indicator function to effectively regularize the neural implicit function learning, towards reliable and high-fidelity surface reconstruction from unorganized points without any normal information. Extensive evaluations on synthetic and real-scan datasets show that our approach consistently outperforms previous approaches, especially when point clouds are incomplete and/or noisy with complex topology structure.
APA
Huang, S., Chen, G., Heng, C.L. & Huang, H.. (2024). NeuralIndicator: Implicit Surface Reconstruction from Neural Indicator Priors. Proceedings of the 41st International Conference on Machine Learning, in Proceedings of Machine Learning Research 235:19622-19634 Available from https://proceedings.mlr.press/v235/huang24b.html.

Related Material