MetaSDF: Meta-learning Signed Distance Functions

Abstract

Neural implicit shape representations are an emerging paradigm that offers many potential benefits over conventional discrete representations, including memory efficiency at a high spatial resolution. Generalizing across shapes with such neural implicit representations amounts to learning priors over the respective function space and enables geometry reconstruction from partial or noisy observations. Existing generalization methods rely on conditioning a neural network on a low-dimensional latent code that is either regressed by an encoder or jointly optimized in the auto-decoder framework. Here, we formalize learning of a shape space as a meta-learning problem and leverage gradient-based meta-learning algorithms to solve this task. We demonstrate that this approach performs on par with auto-decoder based approaches while being an order of magnitude faster at test-time inference. We further demonstrate that the proposed gradient-based method outperforms encoder-decoder based methods that leverage pooling-based set encoders.

Project website: https://vsitzmann.github.io/metasdf/

Cited by in corpus (18)

• ShaRF: Shape-conditioned Radiance Fields from a Single View
• On the Effectiveness of Weight-Encoded Neural Implicit 3D Shapes
• 3D Neural Scene Representations for Visuomotor Control
• Meta-Auto-Decoder for Solving Parametric Partial Differential Equations
• Blended-NeRF: Zero-Shot Object Generation and Blending in Existing Neural Radiance Fields
• Operator Learning with Neural Fields: Tackling PDEs on General Geometries
• Learning Models as Functionals of Signed-Distance Fields for Manipulation Planning
• Multi-NeuS: 3D Head Portraits from Single Image with Neural Implicit Functions
• Signal Transformer: Complex-valued Attention and Meta-Learning for Signal Recognition
• MARF: The Medial Atom Ray Field Object Representation
• Neural Fields in Visual Computing and Beyond
• Meta-Learning 3D Shape Segmentation Functions
• Gradient Origin Networks
• Signed Distance Function Computation from an Implicit Surface
• Geometry encoding for numerical simulations
• Towards Generalising Neural Implicit Representations
• Light Field Implicit Representation for Flexible Resolution Reconstruction
• LatentHuman: Shape-and-Pose Disentangled Latent Representation for Human Bodies