Enhancing the Performance of Convolutional Neural Networks on Quality Degraded Datasets

TL;DR

This paper investigates the impact of noise and quality degradation on convolutional neural networks and proposes a dual-channel architecture to improve classification performance on degraded images.

Contribution

It introduces a novel dual-channel neural network architecture specifically designed to handle quality degraded input images, enhancing robustness.

Findings

Dual-channel model outperforms single models on degraded datasets

The proposed architecture improves classification accuracy under noise and blur

Experimental results demonstrate robustness across various degradation types

Abstract

Despite the appeal of deep neural networks that largely replace the traditional handmade filters, they still suffer from isolated cases that cannot be properly handled only by the training of convolutional filters. Abnormal factors, including real-world noise, blur, or other quality degradations, ruin the output of a neural network. These unexpected problems can produce critical complications, and it is surprising that there has only been minimal research into the effects of noise in the deep neural network model. Therefore, we present an exhaustive investigation into the effect of noise in image classification and suggest a generalized architecture of a dual-channel model to treat quality degraded input images. We compare the proposed dual-channel model with a simple single model and show it improves the overall performance of neural networks on various types of quality degraded input datasets.