Octree Transformer: Autoregressive 3D Shape Generation on Hierarchically Structured Sequences

TL;DR

This paper introduces Octree Transformer, a novel autoregressive model for 3D shape generation that leverages hierarchical octree representations and adaptive compression to efficiently generate complex 3D shapes.

Contribution

It presents a new method to linearize 3D shapes using octrees and an adaptive compression scheme, enabling effective autoregressive modeling with transformers.

Findings

Outperforms state-of-the-art in 3D shape generation

Reduces sequence length significantly with adaptive compression

Enables fully autoregressive sampling of complex shapes

Abstract

Autoregressive models have proven to be very powerful in NLP text generation tasks and lately have gained popularity for image generation as well. However, they have seen limited use for the synthesis of 3D shapes so far. This is mainly due to the lack of a straightforward way to linearize 3D data as well as to scaling problems with the length of the resulting sequences when describing complex shapes. In this work we address both of these problems. We use octrees as a compact hierarchical shape representation that can be sequentialized by traversal ordering. Moreover, we introduce an adaptive compression scheme, that significantly reduces sequence lengths and thus enables their effective generation with a transformer, while still allowing fully autoregressive sampling and parallel training. We demonstrate the performance of our model by comparing against the state-of-the-art in shape generation.