用多头猜测解码加速网格生成，速度提升1.7倍且不损失质量。
XSpecMesh: Quality-Preserving Auto-Regressive Mesh Generation Acceleration via Multi-Head Speculative Decoding
- 通过多头并行猜测多个顶点，单次前向传播预测多个令牌。
- 在保持网格质量的前提下实现1.7倍推理加速。
- 适合需要快速生成高精度3D网格的工业应用或实时系统。
当前自回归模型可生成高质量、拓扑精确的网格；但推理需数千甚至数万次逐令牌预测，导致显著延迟。本文提出XSpecMesh，一种保质加速的自回归网格生成方法。该方法采用轻量级多头猜测解码策略，在单次前向传播中并行预测多个令牌，从而加速推理。进一步提出验证与重采样机制：主干模型验证每个预测令牌，对不满足质量标准的令牌进行重采样。此外，设计了一种蒸馏策略，通过主干模型指导轻量解码头训练，使其预测分布对齐，提升猜测预测成功率。大量实验表明，该方法在不牺牲生成质量的前提下实现了1.7倍的速度提升。代码将开源。
原文摘要 · Abstract (English)
Current auto-regressive models can generate high-quality, topologically precise meshes; however, they necessitate thousands-or even tens of thousands-of next-token predictions during inference, resulting in substantial latency. We introduce XSpecMesh, a quality-preserving acceleration method for auto-regressive mesh generation models. XSpecMesh employs a lightweight, multi-head speculative decoding scheme to predict multiple tokens in parallel within a single forward pass, thereby accelerating inference. We further propose a verification and resampling strategy: the backbone model verifies each predicted token and resamples any tokens that do not meet the quality criteria. In addition, we propose a distillation strategy that trains the lightweight decoding heads by distilling from the backbone model, encouraging their prediction distributions to align and improving the success rate of speculative predictions. Extensive experiments demonstrate that our method achieves a 1.7x speedup without sacrificing generation quality. Our code will be released.
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。