根据3D形状复杂度自适应调整编码粒度，提升生成效率与质量。
Efficient Autoregressive Shape Generation via Octree-Based Adaptive Tokenization
- 基于四次误差的八叉树动态分割，按复杂度分配不同数量的编码单元。
- 相比固定长度编码，令牌数量减少50%且视觉质量相当。
- 适合需要高效高细节3D生成的应用，如游戏资产、数字孪生。
许多3D生成模型依赖变分自编码器（VAEs）学习紧凑的形状表示。然而，现有方法将所有形状编码为固定长度的令牌，忽视了3D数据在尺度和复杂度上的天然差异，导致潜在表示效率低下，影响下游生成效果。本文提出八叉树自适应标记化（Octree-based Adaptive Tokenization），根据形状复杂度动态调整潜在表示维度。该方法基于四次误差的细分准则构建自适应八叉树结构，并通过基于查询的Transformer为每个八叉树单元分配一个形状潜在向量。在此标记化基础上，我们开发了一种基于八叉树的自回归生成模型，有效利用可变尺寸表示进行3D形状生成。大量实验表明，本方法相比固定大小方法令牌数减少50%，同时保持相近的视觉质量；在相似令牌长度下，生成形状质量显著更高；结合下游生成模型后，生成内容在细节和多样性上均优于现有方法。
原文摘要 · Abstract (English)
Many 3D generative models rely on variational autoencoders (VAEs) to learn compact shape representations. However, existing methods encode all shapes into a fixed-size token, disregarding the inherent variations in scale and complexity across 3D data. This leads to inefficient latent representations that can compromise downstream generation. We address this challenge by introducing Octree-based Adaptive Tokenization, a novel framework that adjusts the dimension of latent representations according to shape complexity. Our approach constructs an adaptive octree structure guided by a quadric-error-based subdivision criterion and allocates a shape latent vector to each octree cell using a query-based transformer. Building upon this tokenization, we develop an octree-based autoregressive generative model that effectively leverages these variable-sized representations in shape generation. Extensive experiments demonstrate that our approach reduces token counts by 50% compared to fixed-size methods while maintaining comparable visual quality. When using a similar token length, our method produces significantly higher-quality shapes. When incorporated with our downstream generative model, our method creates more detailed and diverse 3D content than existing approaches.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。