研究人员开发了 ACES（自适应覆盖感知高效采样）框架，旨在提高隐式神经表示（INRs）的训练效率。与先前逐点或均匀采样的旧方法不同，ACES 使用自适应空间分区来确保域覆盖，并优先处理信息丰富的区域进行训练。实验表明，这种方法可以降低梯度方差并提高优化效率，从而加快收敛速度并降低错误率，尤其是在处理具有局部复杂性的场时。 AI
影响 该框架有望加速用于科学模拟和数据建模的复杂模型的训练。
排序理由 该集群描述了一篇关于用于神经网络训练的新型采样框架的学术论文。
在 Hugging Face Daily Papers 阅读 →
AI 生成摘要 · Google Gemini · 来自 2 个来源。 我们如何撰写摘要 →
完整方法见我们的编辑标准。
arXiv:2610.02410v1 Announce Type: cross Abstract: Implicit neural representations (INRs) provide a flexible framework for modeling high-dimensional continuous fields, but their training is often inefficient due to uniform subsampling that ignores spatial heterogeneity. Existing a…
Implicit neural representations (INRs) provide a flexible framework for modeling high-dimensional continuous fields, but their training is often inefficient due to uniform subsampling that ignores spatial heterogeneity. Existing adaptive sampling methods partially address this is…