BriefGPT.xyz
Oct, 2024
增强深度学习的残差科尔莫戈罗夫-阿诺德网络
Residual Kolmogorov-Arnold Network for Enhanced Deep Learning
HTML
PDF
Ray Congrui Yu, Sherry Wu, Jiang Gui
TL;DR
本研究解决了卷积神经网络在深层网络中难以有效捕捉长距离复杂非线性依赖的问题。通过在CNN框架中引入残差科尔莫戈罗夫-阿诺德网络(RKAN),并利用切比雪夫多项式作为KAN卷积的基础,我们实现了更具表达性和适应性的特征表示。研究结果表明,RKAN能够在视觉数据分析中增强深度CNN的能力。
Abstract
Despite the strong performance in many
Computer Vision
tasks, Convolutional Neural Networks (CNNs) can sometimes struggle to efficiently capture long-range, complex
Non-linear Dependencies
in deeper layers of the
→