Dense-Sparse Deep Convolutional Neural Networks Training for Image Denoising

TL;DR

This paper introduces a dense-sparse-dense training method for deep convolutional neural networks that achieves comparable image denoising performance with fewer parameters and faster processing times.

Contribution

It proposes an enhanced training procedure that reduces the number of trainable parameters while maintaining denoising effectiveness, improving efficiency over existing deep CNN methods.

Findings

Achieves similar denoising performance with fewer parameters.

Reduces training and processing time significantly.

Demonstrates effectiveness of dense-sparse-dense training in image denoising.

Abstract

Recently, deep learning methods such as the convolutional neural networks have gained prominence in the area of image denoising. This is owing to their proven ability to surpass state-of-the-art classical image denoising algorithms such as block-matching and 3D filtering algorithm. Deep denoising convolutional neural networks use many feed-forward convolution layers with added regularization methods of batch normalization and residual learning to speed up training and improve denoising performance significantly. However, this comes at the expense of a huge number of trainable parameters. In this paper, we show that by employing an enhanced dense-sparse-dense network training procedure to the deep denoising convolutional neural networks, comparable denoising performance level can be achieved at a significantly reduced number of trainable parameters. We derive motivation from the fact that networks trained using the dense-sparse-dense approach have been shown to attain performance boost with reduced number of parameters. The proposed reduced deep denoising convolutional neural networks network is an efficient denoising model with significantly reduced parameters and comparable performance to the deep denoising convolutional neural networks. Additionally, denoising was achieved at significantly reduced processing time.