Rectifier Neural Network with a Dual-Pathway Architecture for Image Denoising

TL;DR

This paper introduces a dual-pathway rectifier neural network for image denoising, replacing tanh with rectifier functions, and demonstrates improved performance and speed, especially with low noise levels.

Contribution

The paper proposes a novel dual-pathway rectifier neural network architecture for image denoising, replacing tanh activation and showing superior results.

Findings

Achieves better denoising performance than tanh-based models.

Faster convergence especially with low noise levels.

Demonstrates the effectiveness of rectifier functions in image denoising.

Abstract

Recently deep neural networks based on tanh activation function have shown their impressive power in image denoising. In this letter, we try to use rectifier function instead of tanh and propose a dual-pathway rectifier neural network by combining two rectifier neurons with reversed input and output weights in the same hidden layer. We drive the equivalent activation function and compare it to some typical activation functions for image denoising under the same network architecture. The experimental results show that our model achieves superior performances faster especially when the noise is small.