发表机构
PUC-Rio; Universidade Federal de Santa Maria; Google DeepMind; University of Coimbra; IMPA（里约热内卢天主教大学; 圣玛丽亚联邦大学; 谷歌DeepMind; 科英布拉大学; 国家纯数学与应用数学研究所）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
M-plicits提出一种嵌套多尺度残差MLP框架，通过窄带监督和解析法线计算，实现高效、鲁棒的神经隐式曲面重建与实时渲染，在多个基准上取得最优性能。
AI 中文摘要
将输入坐标用正弦函数编码进多层感知器（MLPs），已被证明对于定义为零水平集的曲面的隐式神经表示（INRs）是有效的。然而，现有方法往往难以平衡训练效率、渲染速度和噪声鲁棒性：单MLP方法在推理时成本高昂，基于网格的表示速度快但可能限制曲面平滑度并过拟合输入噪声，而先前的多尺度方法由于硬频谱截断经常捕获噪声并产生伪影。为解决这些局限性，我们提出M-plicits，一个多尺度框架，将曲面建模为通过一系列嵌套邻域训练的MLPs的残差和。与依赖标准全域采样并需要昂贵的网格提取进行可视化的现有残差方法不同，我们的方法将监督严格限制在先前零水平集周围的窄带上。这种嵌套设计自然地对噪声输入数据提供鲁棒性：粗网络充当低通滤波器，建立干净的几何先验，而后续残差逐步细化几何而不拟合高频伪影。我们进一步引入多尺度球追踪算法和基于GEMM的解析法线计算，完全绕过自动微分，实现高保真实时渲染。在Stanford和Thingi32上，M-plicits在粗配置中达到最佳平均Chamfer距离，在细配置中达到最佳中位数Chamfer距离和IoU，其噪声鲁棒性显著优于iNGP、BACON和IDF，同时使用的参数比基于网格的基线少一个数量级。代码、模型和数据将在该https URL发布。
英文摘要
Encoding input coordinates with sinusoidal functions into multi-layer perceptrons (MLPs) has proven effective for implicit neural representations (INRs) of surfaces defined as zero-level sets. However, existing methods often struggle to balance training efficiency, rendering speed, and noise robustness: single-MLP approaches are expensive at inference, grid-based representations are fast but can limit surface smoothness and overfit input noise, and previous multiscale approaches frequently capture noise and produce artifacts due to hard spectral truncation. To address these limitations, we propose M-plicits, a multiscale framework that models surfaces as a residual sum of MLPs trained via a sequence of nested neighborhoods. Unlike existing residual approaches that rely on standard domain-wide sampling and require costly mesh extraction for visualization, our method strictly localizes supervision to narrow bands around the previous zero-level sets. This nested design naturally provides robustness against noisy input data: the coarse network acts as a low-pass filter that establishes a clean geometric prior, while subsequent residuals progressively refine the geometry without fitting to high-frequency artifacts. We further introduce a multiscale sphere-tracing algorithm and a GEMM-based analytical normal computation that bypasses auto-differentiation entirely, yielding high-fidelity real-time rendering. On Stanford and Thingi32, M-plicits achieves the best mean Chamfer distance in the coarse configuration and the best median Chamfer distance and IoU in the fine configuration, with substantially better noise robustness than iNGP, BACON, and IDF, while using an order of magnitude fewer parameters than grid-based baselines. Code, models, and data are available at https://github.com/dsilvavinicius/m-plicits.
CommentsAccepted at NeurIPS 2026 (poster). Project page: https://dsilvavinicius.github.io/m-plicits/ - code, models and data: https://github.com/dsilvavinicius/m-plicits
