让神经渲染更高效精准，融入物理光照原理
RenderFormer++: Scalable and Physics-Informed Feed-Forward Neural Rendering
- 用物理光照规律引导注意力机制，提升渲染真实性
- 对象级特征聚合使计算量大幅降低，支持大规模场景
- 适合需要快速高质量全局光照的3D渲染研究者
我们提出 RenderFormer++，一种可扩展且融合物理先验的前馈神经渲染框架，用于网格场景中的全局光照。现有基于 Transformer 的神经渲染方法如 RenderFormer 虽具良好跨场景泛化能力，但缺乏显式光传输先验，且因三角面片级注意力导致扩展性差。为此，我们引入物理信息传输引导（PITG），将渲染方程启发的归纳偏置嵌入注意力机制，并设计传输一致性损失，促进物理合理的光传输建模。同时提出分层对象中心标记化（HOCT），通过可学习查询与交叉注意力将三角面片特征聚合为紧凑的对象级标记，显著降低计算与内存开销。大量实验表明，RenderFormer++ 在复杂大场景中实现可扩展、强泛化的前馈全局光照渲染，质量媲美且效率远超 RenderFormer。代码将在录用后公开。
原文摘要 · Abstract (English)
We present RenderFormer++, a scalable and physics-informed feed-forward neural rendering framework for global illumination in mesh scenes. Existing Transformer-based neural rendering methods such as RenderFormer achieve promising cross-scene generalization, but lack explicit transport priors and scale poorly due to quadratic triangle-level attention. To address these issues, we introduce Physics-Informed Transport Guidance (PITG), which embeds rendering-equation-inspired inductive biases into the attention mechanism and introduces a transport consistency loss, encouraging physics-informed light transport modeling. We further propose Hierarchical Object-Centric Tokenization (HOCT), which aggregates triangle-level features into compact object-level tokens via cross-attention with learnable queries, substantially reducing computational and memory costs. Extensive experiments demonstrate that RenderFormer++ achieves scalable and generalizable feed-forward global illumination rendering across complex large-scale scenes with competitive rendering quality and substantially improved efficiency over RenderFormer. The code will be made publicly available upon acceptance.
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。