Shape As Points: A Differentiable Poisson Solver

Authors: Songyou Peng, Chiyu Jiang, Yiyi Liao, Michael Niemeyer, Marc Pollefeys, Andreas Geiger

NeurIPS 2021 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental To investigate the aforementioned properties, we perform a set of controlled experiments. Moreover, we demonstrate state-of-the-art performance in reconstructing surface geometry from unoriented point clouds in two settings: an optimization-based setting that does not require training and is applicable to a wide range of shapes, and a learning-based setting for conditional shape reconstruction that is robust to noisy point clouds and outliers.
Researcher Affiliation Collaboration Songyou Peng1,2 Chiyu Max Jiang Yiyi Liao2,3 Michael Niemeyer2,3 Marc Pollefeys1,4 Andreas Geiger2,3 1ETH Zurich 2Max Planck Institute for Intelligent Systems, Tübingen 3University of Tübingen 4Microsoft
Pseudocode No The paper does not contain any structured pseudocode or algorithm blocks.
Open Source Code Yes Code is available at https://github.com/autonomousvision/shape_as_points.
Open Datasets Yes Datasets: We use the following datasets for optimization-based reconstruction: 1) Thingi10K [71], 2) Surface reconstruction benchmark (SRB) [65] and 3) D-FAUST [6]. Similar to prior works, we use 5 objects per dataset [19,23,65]. For learning-based object-level reconstruction, we consider all 13 classes of the Shape Net [9] subset, using the train/val/test split from [11].
Dataset Splits Yes For learning-based object-level reconstruction, we consider all 13 classes of the Shape Net [9] subset, using the train/val/test split from [11].
Hardware Specification Yes On average, our method requires 12 ms for computing a 1283 indicator grid from 15K points on a single NVIDIA GTX 1080Ti GPU. Computing a 2563 indicator grid requires 140 ms. Optimization time is evaluated on a single GTX 1080Ti GPU for IGR, Point2Mesh and our method.
Software Dependencies No The paper states 'We implement all models in Py Torch [51] and use the Adam optimizer [30]' and refers to 'scikit-image [60]', but it does not provide specific version numbers for these software components.
Experiment Setup Yes We implement all models in Py Torch [51] and use the Adam optimizer [30] with a learning rate of 5e-4. We choose k = 7 for all learning-based reconstruction experiments. More specifically, we start optimizing at an indicator grid resolution of 323 for 1000 iterations, from which we obtain a coarse shape. Next, we sample from this coarse mesh and continue optimization at a resolution of 643 for 1000 iterations. We repeat this process until we reach the target resolution (2563) at which we acquire the final output mesh.