MatE：基于几何先验的单图像材质提取
MatE: Material Extraction from Single-Image via Geometric Prior
用估计深度图作为几何先验做粗矫正，再用双分支扩散模型把单张图转成可平铺的PBR材质贴图集（albedo/normal/roughness/height），其旋转与尺度对齐训练数据带来的等变性思路可迁移到其他需要消除透视与光照干扰的逆渲染任务。
原文依据 MatE first performs coarse rectification using an estimated depth map as a geometric prior, and then employs a dual-branch diffusion model.
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论创建高保真、基于物理的渲染（PBR）材质仍是许多图形流水线中的瓶颈，通常需要专用设备和专家驱动的后处理。为降低这一过程的门槛，我们提出 MatE，一种在不受约束的真实世界条件下从单张图像生成可平铺 PBR 材质的新方法。给定一张图像和用户提供的掩码，MatE 首先利用估计的深度图作为几何先验进行粗矫正，然后采用双分支扩散模型。借助从旋转对齐与尺度对齐训练数据中学到的一致性，该模型进一步矫正粗结果中的残余畸变，并将其转换为完整的材质贴图集，包括反照率、法线、粗糙度和高度。我们的框架对输入图像未知的光照与透视具有不变性，从而能够从随手拍摄中恢复本征材质属性。在合成数据与真实数据上的全面实验证明了我们方法的有效性与鲁棒性，使用户能够从真实世界图像创建逼真的材质。
英文原文摘要 1195 字符
The creation of high-fidelity, physically-based rendering (PBR) materials remains a bottleneck in many graphics pipelines, typically requiring specialized equipment and expert-driven post-processing. To democratize this process, we present MatE, a novel method for generating tileable PBR materials from a single image taken under unconstrained, real-world conditions. Given an image and a user-provided mask, MatE first performs coarse rectification using an estimated depth map as a geometric prior, and then employs a dual-branch diffusion model. Leveraging a learned consistency from rotation-aligned and scale-aligned training data, this model further rectifies residual distortions from the coarse result and translates it into a complete set of material maps, including albedo, normal, roughness and height. Our framework achieves invariance to the unknown illumination and perspective of the input image, allowing for the recovery of intrinsic material properties from casual captures. Through comprehensive experiments on both synthetic and real-world data, we demonstrate the efficacy and robustness of our approach, enabling users to create realistic materials from real-world images.