Deep Level Sets: Implicit Surface Representations for 3D Shape Inference

Abstract

Existing 3D surface representation approaches are unable to accurately classify pixels and their orientation lying on the boundary of an object. Thus resulting in coarse representations which usually require post-processing steps to extract 3D surface meshes. To overcome this limitation, we propose an end-to-end trainable model that directly predicts implicit surface representations of arbitrary topology by optimising a novel geometric loss function. Specifically, we propose to represent the output as an oriented level set of a continuous embedding function, and incorporate this in a deep end-to-end learning framework by introducing a variational shape inference formulation. We investigate the benefits of our approach on the task of 3D surface prediction and demonstrate its ability to produce a more accurate reconstruction compared to voxel-based representations. We further show that our model is flexible and can be applied to a variety of shape inference problems.

References in corpus (5)

• MobileNetV2: Inverted Residuals and Linear Bottlenecks
• Unsupervised Learning of 3D Structure from Images
• Robust Watertight Manifold Surface Generation Method for ShapeNet Models
• Matryoshka Networks: Predicting 3D Geometry via Nested Shape Layers
• A Point Set Generation Network for 3D Object Reconstruction from a Single Image

Cited by in corpus (23)

• H-NeRF: Neural Radiance Fields for Rendering and Temporal Reconstruction of Humans in Motion
• Neural-Pull: Learning Signed Distance Functions from Point Clouds by Learning to Pull Space onto Surfaces
• Implicit Geometric Regularization for Learning Shapes
• Neural Unsigned Distance Fields for Implicit Function Learning
• Cascaded Refinement Network for Point Cloud Completion
• DIST: Rendering Deep Implicit Signed Distance Function with Differentiable Sphere Tracing
• Deep Mesh Reconstruction from Single RGB Images via Topology Modification Networks
• Higher-Order Function Networks for Learning Composable 3D Object Representations
• Grasping Field: Learning Implicit Representations for Human Grasps
• SDFDiff: Differentiable Rendering of Signed Distance Fields for 3D Shape Optimization
• Analytic Marching: An Analytic Meshing Solution from Deep Implicit Surface Networks
• Learning Gradient Fields for Shape Generation
• Total3DUnderstanding: Joint Layout, Object Pose and Mesh Reconstruction for Indoor Scenes from a Single Image
• LightSAL: Lightweight Sign Agnostic Learning for Implicit Surface Representation
• A Simple and Scalable Shape Representation for 3D Reconstruction
• SkeletonNet: A Topology-Preserving Solution for Learning Mesh Reconstruction of Object Surfaces from RGB Images
• Neural Mesh Flow: 3D Manifold Mesh Generation via Diffeomorphic Flows
• Point Cloud Completion by Learning Shape Priors
• SeqXY2SeqZ: Structure Learning for 3D Shapes by Sequentially Predicting 1D Occupancy Segments From 2D Coordinates
• Surface HOF: Surface Reconstruction from a Single Image Using Higher Order Function Networks
• Cascaded Refinement Network for Point Cloud Completion with Self-supervision
• Towards Generalising Neural Implicit Representations
• LatentHuman: Shape-and-Pose Disentangled Latent Representation for Human Bodies