N-fold Superposition: Improving Neural Networks by Reducing the Noise in Feature Maps

TL;DR

This paper introduces N-fold Superposition, a technique that reduces noise in feature maps to enhance CNN performance, accelerate convergence, and prevent overfitting with minimal additional parameters.

Contribution

The paper proposes a novel superposition method to improve feature map coupling in CNNs, leading to faster convergence and better classification accuracy.

Findings

Increases convergence speed of neural networks.

Improves classification performance.

Widely widens the loss function's minima range.

Abstract

Considering the use of Fully Connected (FC) layer limits the performance of Convolutional Neural Networks (CNNs), this paper develops a method to improve the coupling between the convolution layer and the FC layer by reducing the noise in Feature Maps (FMs). Our approach is divided into three steps. Firstly, we separate all the FMs into n blocks equally. Then, the weighted summation of FMs at the same position in all blocks constitutes a new block of FMs. Finally, we replicate this new block into n copies and concatenate them as the input to the FC layer. This sharing of FMs could reduce the noise in them apparently and avert the impact by a particular FM on the specific part weight of hidden layers, hence preventing the network from overfitting to some extent. Using the Fermat Lemma, we prove that this method could make the global minima value range of the loss function wider, by which makes it easier for neural networks to converge and accelerates the convergence process. This method does not significantly increase the amounts of network parameters (only a few more coefficients added), and the experiments demonstrate that this method could increase the convergence speed and improve the classification performance of neural networks.