arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2407.07655cs.LG
收藏

选择性G-双谱及其反演:在G-不变网络中的应用

The Selective G-Bispectrum and its Inversion: Applications to G-Invariant Networks

ICTEAM, UCLouvain(天主教鲁汶大学 ICTEAM学院);Atmo(Atmo公司);UC Santa Barbara(加州大学圣巴巴拉分校)

Simon Mataigne, Johan Mathe, Sophia Sanborn, Christopher Hillar, Nina Miolane

AI总结:

针对G-双谱计算成本高的问题,提出选择性G-双谱,将复杂度从O(|G|^2)降至O(|G|),并证明其数学性质及在G-不变网络中的准确性和鲁棒性提升。

摘要:

信号处理与深度学习中的一个重要问题是实现对任务无关的干扰因素的“不变性”。由于许多此类因素可描述为群$G$的作用(例如旋转、平移、缩放),我们希望方法具有$G$-不变性。$G$-双谱提取给定信号在群作用下的所有特征:例如,图像中物体的形状,但不包括其方向。因此,$G$-双谱已被纳入深度神经网络架构,作为实现$G$-不变性的计算原语——类似于池化机制,但具有更高的选择性和鲁棒性。然而,$G$-双谱的计算成本($\mathcal{O}(|G|^2)$,其中$|G|$为群的大小)限制了其广泛应用。在此,我们证明$G$-双谱计算中存在冗余,可将其简化为具有$\mathcal{O}(|G|)$复杂度的“选择性$G$-双谱”。我们证明了选择性$G$-双谱的优良数学性质,并展示了其在神经网络中的集成相比传统方法如何提升准确性和鲁棒性,同时相比完整$G$-双谱享有显著的加速。

英文摘要:

An important problem in signal processing and deep learning is to achieve \textit{invariance} to nuisance factors not relevant for the task. Since many of these factors are describable as the action of a group $G$ (e.g. rotations, translations, scalings), we want methods to be $G$-invariant. The $G$-Bispectrum extracts every characteristic of a given signal up to group action: for example, the shape of an object in an image, but not its orientation. Consequently, the $G$-Bispectrum has been incorporated into deep neural network architectures as a computational primitive for $G$-invariance\textemdash akin to a pooling mechanism, but with greater selectivity and robustness. However, the computational cost of the $G$-Bispectrum ($\mathcal{O}(|G|^2)$, with $|G|$ the size of the group) has limited its widespread adoption. Here, we show that the $G$-Bispectrum computation contains redundancies that can be reduced into a \textit{selective $G$-Bispectrum} with $\mathcal{O}(|G|)$ complexity. We prove desirable mathematical properties of the selective $G$-Bispectrum and demonstrate how its integration in neural networks enhances accuracy and robustness compared to traditional approaches, while enjoying considerable speeds-up compared to the full $G$-Bispectrum.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑