Neural-Pull: Learning Signed Distance Functions from Point Clouds by Learning to Pull Space onto Surfaces

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 \textit{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.

arXiv commentTo appear at ICML2021. Code and data are available at https://github.com/mabaorui/NeuralPull

References in corpus (8)

• Learning to Infer Implicit Surfaces without 3D Supervision
• Deep Level Sets: Implicit Surface Representations for 3D Shape Inference
• Point2SpatialCapsule: Aggregating Features and Spatial Relationships of Local Regions on Point Clouds using Spatial-aware Capsules
• Neural Geometric Level of Detail: Real-time Rendering with Implicit 3D Shapes
• DRWR: A Differentiable Renderer without Rendering for Unsupervised 3D Structure Learning from Silhouette Images
• Learning Occupancy Function from Point Clouds for Surface Reconstruction
• Generative Models as Distributions of Functions
• ACORN: Adaptive Coordinate Networks for Neural Scene Representation

Cited by in corpus (4)

• Fast Learning Radiance Fields by Shooting Much Fewer Rays
• Neural-IMLS: Self-supervised Implicit Moving Least-Squares Network for Surface Reconstruction
• PG-NeuS: Robust and Efficient Point Guidance for Multi-View Neural Surface Reconstruction
• SnowflakeNet: Point Cloud Completion by Snowflake Point Deconvolution with Skip-Transformer