通过并行渐进预测生成多尺度高斯点云，提升3D内容生成效率与质量。
ARGS: Auto-Regressive Gaussian Splatting via Parallel Progressive Next-Scale Prediction
- 利用层级树结构实现并行渐进式生成，仅需O(log n)步完成
- 通过反向简化策略引导生成，保持细节一致性，支持可控细节层次
- 基于树形Transformer增强结构连贯性，适合高质量3D生成任务
自回归框架在2D图像的多尺度渐进预测中展现出生成多样且复杂内容的强大潜力。然而，将该范式扩展至3D物体生成仍处于空白状态。本文提出自回归高斯点云（ARGS），一种基于细节层级并行进行下一级预测的生成框架。我们设计了一种高斯简化策略，并反向执行简化过程以指导下一尺度生成。得益于分层树结构，生成过程仅需$/mathcal{O}( ext{log } n)$步，其中$n$为点数。此外，我们引入基于树结构的Transformer模型，实现自回归预测树形结构，使叶节点可关注其内部祖先，从而增强结构一致性。大量实验表明，该方法能有效生成具有可控细节层次、高视觉保真度且计算开销合理的多尺度高斯表示。
原文摘要 · Abstract (English)
Auto-regressive frameworks for next-scale prediction of 2D images have demonstrated strong potential for producing diverse and sophisticated content by progressively refining a coarse input. However, extending this paradigm to 3D object generation remains largely unexplored. In this paper, we introduce auto-regressive Gaussian splatting (ARGS), a framework for making next-scale predictions in parallel for generation according to levels of detail. We propose a Gaussian simplification strategy and reverse the simplification to guide next-scale generation. Benefiting from the use of hierarchical trees, the generation process requires only \(\mathcal{O}(\log n)\) steps, where \(n\) is the number of points. Furthermore, we propose a tree-based transformer to predict the tree structure auto-regressively, allowing leaf nodes to attend to their internal ancestors to enhance structural consistency. Extensive experiments demonstrate that our approach effectively generates multi-scale Gaussian representations with controllable levels of detail, visual fidelity, and a manageable time consumption budget.
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。