用双曲正弦激活函数提升隐式神经表示的频率建模能力
H-SIREN: Improving implicit neural representations with hyperbolic periodic functions
- 首层激活函数由sin(x)改为sin(sinh(2x))，增强高频建模能力
- 在视觉与流体模拟任务中性能超越多个最新隐式神经表示方法
- 适合需要高精度细节建模的计算机视觉与物理仿真场景
隐式神经表示（INR）已被广泛应用于计算机视觉到物理模拟等任务，通过求解偏微分方程实现。现有基于多层感知机的INR常以正弦激活函数为基准，但研究表明其支持的频率范围有限，且易导致解过平滑。本文提出H-SIREN，将第一层激活函数从sin(x)替换为sin(sinh(2x))，有效缓解该问题。实验在多种计算机视觉与流体流动任务中验证了H-SIREN的优越性，显著优于多个最先进的INR模型。
原文摘要 · Abstract (English)
Implicit neural representations (INR) have been recently adopted in various applications ranging from computer vision tasks to physics simulations by solving partial differential equations. Among existing INR-based works, multi-layer perceptrons with sinusoidal activation functions find widespread applications and are also frequently treated as a baseline for the development of better activation functions for INR applications. Recent investigations claim that the use of sinusoidal activation functions could be sub-optimal due to their limited supported frequency set as well as their tendency to generate over-smoothed solutions. We provide a simple solution to mitigate such an issue by changing the activation function at the first layer from $\sin(x)$ to $\sin(\sinh(2x))$. We demonstrate H-SIREN in various computer vision and fluid flow problems, where it surpasses the performance of several state-of-the-art INRs.
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。