FastMesh: Efficient Artistic Mesh Generation via Component Decoupling

TL;DR

FastMesh introduces a novel mesh generation framework that separates vertex and face generation, significantly reducing token redundancy and enabling faster, higher-quality artistic mesh creation.

Contribution

The paper presents a new method that decouples vertex and face generation, employing autoregressive and bidirectional models to improve efficiency and quality in mesh synthesis.

Findings

Achieves over 8x faster mesh generation speed.

Reduces token count to 23% of previous methods.

Produces higher quality meshes with natural vertex arrangements.

Abstract

Recent mesh generation approaches typically tokenize triangle meshes into sequences of tokens and train autoregressive models to generate these tokens sequentially. Despite substantial progress, such token sequences inevitably reuse vertices multiple times to fully represent manifold meshes, as each vertex is shared by multiple faces. This redundancy leads to excessively long token sequences and inefficient generation processes. In this paper, we propose an efficient framework that generates artistic meshes by treating vertices and faces separately, significantly reducing redundancy. We employ an autoregressive model solely for vertex generation, decreasing the token count to approximately 23% of that required by the most compact existing tokenizer. Next, we leverage a bidirectional transformer to complete the mesh in a single step by capturing inter-vertex relationships and constructing the adjacency matrix that defines the mesh faces. To further improve the generation quality, we introduce a fidelity enhancer to refine vertex positioning into more natural arrangements and propose a post-processing framework to remove undesirable edge connections. Experimental results show that our method achieves more than 8x faster speed on mesh generation compared to state-of-the-art approaches, while producing higher mesh quality.