Mesh-Pro: 异步优势引导的排名偏好优化用于艺术家风格的四边形网格生成
Mesh-Pro: Asynchronous Advantage-guided Ranking Preference Optimization for Artist-style Quadrilateral Mesh Generation
AI总结：
Mesh-Pro通过异步优势引导的排名偏好优化提升3D网格生成的训练效率和生成质量，引入了新的标记化和奖励机制，实现了艺术家风格网格的先进性能。
摘要：
强化学习（RL）在文本和图像生成中表现出显著的成功，但其在3D生成中的潜力仍 largely未被探索。现有尝试通常依赖于离线直接偏好优化（DPO）方法，这导致训练效率低且泛化能力有限。在本工作中，我们旨在提高RL在3D网格生成中的训练效率和生成质量。具体而言，（1）我们设计了首个异步在线RL框架，专门针对3D网格生成后的训练效率提升，其速度比同步RL快3.75倍。（2）我们提出了优势引导的排名偏好优化（ARPO），一种新的RL算法，比目前用于3D网格生成的RL算法（如DPO和组相对策略优化（GRPO））在训练效率和泛化能力之间取得了更好的平衡。（3）基于异步ARPO，我们提出了Mesh-Pro，其还引入了新的对角线感知混合三角形-四边形标记化用于网格表示和基于射线的奖励用于几何完整性。Mesh-Pro在艺术和密集网格上实现了最先进的性能。
英文摘要：
Reinforcement learning (RL) has demonstrated remarkable success in text and image generation, yet its potential in 3D generation remains largely unexplored. Existing attempts typically rely on offline direct preference optimization (DPO) method, which suffers from low training efficiency and limited generalization. In this work, we aim to enhance both the training efficiency and generation quality of RL in 3D mesh generation. Specifically, (1) we design the first asynchronous online RL framework tailored for 3D mesh generation post-training efficiency improvement, which is 3.75$\times$ faster than synchronous RL. (2) We propose Advantage-guided Ranking Preference Optimization (ARPO), a novel RL algorithm that achieves a better trade-off between training efficiency and generalization than current RL algorithms designed for 3D mesh generation, such as DPO and group relative policy optimization (GRPO). (3) Based on asynchronous ARPO, we propose Mesh-Pro, which additionally introduces a novel diagonal-aware mixed triangular-quadrilateral tokenization for mesh representation and a ray-based reward for geometric integrity. Mesh-Pro achieves state-of-the-art performance on artistic and dense meshes.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
