GRASS: Generative Recursive Autoencoders for Shape Structures

TL;DR

This paper presents GRASS, a recursive autoencoder architecture that captures hierarchical structures of 3D shapes for encoding, synthesis, and generation, enabling diverse shape modeling without supervision.

Contribution

The paper introduces a novel recursive neural network autoencoder for hierarchical 3D shape structures, including a generative adversarial component for plausible shape synthesis.

Findings

Learned meaningful structural hierarchies without supervision

Produced compact codes for shape classification and matching

Enabled shape synthesis and interpolation with topology variations

Abstract

We introduce a novel neural network architecture for encoding and synthesis of 3D shapes, particularly their structures. Our key insight is that 3D shapes are effectively characterized by their hierarchical organization of parts, which reflects fundamental intra-shape relationships such as adjacency and symmetry. We develop a recursive neural net (RvNN) based autoencoder to map a flat, unlabeled, arbitrary part layout to a compact code. The code effectively captures hierarchical structures of man-made 3D objects of varying structural complexities despite being fixed-dimensional: an associated decoder maps a code back to a full hierarchy. The learned bidirectional mapping is further tuned using an adversarial setup to yield a generative model of plausible structures, from which novel structures can be sampled. Finally, our structure synthesis framework is augmented by a second trained module that produces fine-grained part geometry, conditioned on global and local structural context, leading to a full generative pipeline for 3D shapes. We demonstrate that without supervision, our network learns meaningful structural hierarchies adhering to perceptual grouping principles, produces compact codes which enable applications such as shape classification and partial matching, and supports shape synthesis and interpolation with significant variations in topology and geometry.