Initialization Using Perlin Noise for Training Networks with a Limited Amount of Data
Initialization Using Perlin Noise for Training Networks with a Limited Amount of Data
复制标题
DOI：
10.1109/icpr48806.2021.9412955
复制
发表时间：
2021-01
期刊：
影响因子：
--
通讯作者：
Nakamasa Inoue;Eisuke Yamagata;Hirokatsu Kataoka
中科院分区：
文献类型：
--
作者：
Nakamasa Inoue;Eisuke Yamagata;Hirokatsu Kataoka
本文提出了一种新的基于柏林噪声的网络初始化方法，用于训练数据量有限的图像分类网络。我们的主要思想是通过解决一个人工噪声分类问题来初始化网络参数，目的是将柏林噪声样本分类到它们的噪声类别中。具体来说，该方法包括两个步骤。首先，它生成带有基于噪声复杂度定义的类别标签的柏林噪声样本。其次，解决了一个分类问题，优化网络参数对生成的噪声样本进行分类。该方法为图像分类生成一组合理的初始权重（滤波器）。据我们所知，这是第一次通过解决人工优化问题来初始化网络，而不使用任何现实世界的图像。实验表明，该方法在四种图像分类数据集上优于传统的初始化方法。
We propose a novel network initialization method using Perlin noise for training image classification networks with a limited amount of data. Our main idea is to initialize the network parameters by solving an artificial noise classification problem, where the aim is to classify Perlin noise samples into their noise categories. Specifically, the proposed method consists of two steps. First, it generates Perlin noise samples with category labels defined based on noise complexity. Second, it solves a classification problem, in which network parameters are optimized to classify the generated noise samples. This method produces a reasonable set of initial weights (filters) for image classification. To the best of our knowledge, this is the first work to initialize networks by solving an artificial optimization problem without using any real-world images. Our experiments show that the proposed method outperforms conventional initialization methods on four image classification datasets.