Abstract:A signed distance function (SDF) as the 3D shape description is one of the most effective approaches to represent 3D geometry for rendering and reconstruction. Our work is inspired by the state-of-the-art method DeepSDF [17] that learns and analyzes the 3D shape as the iso-surface of its shell and this method has shown promising results especially in the 3D shape reconstruction and compression domain. In this paper, we consider the degeneration problem of reconstruction coming from the capacity decrease of the DeepSDF model, which approximates the SDF with a neural network and a single latent code. We propose Local Geometry Code Learning (LGCL), a model that improves the original DeepSDF results by learning from a local shape geometry of the full 3D shape. We add an extra graph neural network to split the single transmittable latent code into a set of local latent codes distributed on the 3D shape. Mentioned latent codes are used to approximate the SDF in their local regions, which will alleviate the complexity of the approximation compared to the original DeepSDF. Furthermore, we introduce a new geometric loss function to facilitate the training of these local latent codes. Note that other local shape adjusting methods use the 3D voxel representation, which in turn is a problem highly difficult to solve or even is insolvable. In contrast, our architecture is based on graph processing implicitly and performs the learning regression process directly in the latent code space, thus make the proposed architecture more flexible and also simple for realization. Our experiments on 3D shape reconstruction demonstrate that our LGCL method can keep more details with a significantly smaller size of the SDF decoder and outperforms considerably the original DeepSDF method under the most important quantitative metrics.

A-SDF: Learning Disentangled Signed Distance Functions for Articulated Shape Representation

DeepSDF: Learning Continuous Signed Distance Functions for Shape Representation

LatentHuman: Shape-and-Pose Disentangled Latent Representation for Human Bodies

RobotSDF: Implicit Morphology Modeling for the Robotic Arm

gSDF: Geometry-Driven Signed Distance Functions for 3D Hand-Object Reconstruction

A Deep Signed Directional Distance Function for Object Shape Representation

AlignSDF: Pose-Aligned Signed Distance Fields for Hand-Object Reconstruction

Few-Shot Unsupervised Implicit Neural Shape Representation Learning with Spatial Adversaries

Learning Part Motion of Articulated Objects Using Spatially Continuous Neural Implicit Representations

SDF-SRN: Learning Signed Distance 3D Object Reconstruction from Static Images

Learning Implicit Representation for Reconstructing Articulated Objects

Shared Latent Membership Enables Joint Shape Abstraction and Segmentation With Deformable Superquadrics

Unsigned Orthogonal Distance Fields: An Accurate Neural Implicit Representation for Diverse 3D Shapes

Deep Part Induction from Articulated Object Pairs

Self-supervised Learning of Implicit Shape Representation with Dense Correspondence for Deformable Objects

Shallow Signed Distance Functions for Kinematic Collision Bodies

Diffusion-SDF: Conditional Generative Modeling of Signed Distance Functions

PoseSDF: Simultaneous 3D Human Shape Reconstruction and Gait Pose Estimation Using Signed Distance Functions

Building Rearticulable Models for Arbitrary 3D Objects from 4D Point Clouds

3D Shapes Local Geometry Codes Learning with SDF

Latent Partition Implicit with Surface Codes for 3D Representation