SoftReMish: A Novel Activation Function for Enhanced Convolutional Neural Networks for Visual Recognition Performance

TL;DR

SoftReMish is a new activation function that improves CNN performance in image classification by achieving better convergence and higher accuracy than existing functions like ReLU, Tanh, and Mish.

Contribution

The paper introduces SoftReMish, a novel activation function that enhances CNN training efficiency and accuracy for visual recognition tasks.

Findings

SoftReMish outperforms ReLU, Tanh, and Mish in training loss and accuracy.

Achieves 99.41% validation accuracy on MNIST.

Demonstrates improved convergence behavior.

Abstract

In this study, SoftReMish, a new activation function designed to improve the performance of convolutional neural networks (CNNs) in image classification tasks, is proposed. Using the MNIST dataset, a standard CNN architecture consisting of two convolutional layers, max pooling, and fully connected layers was implemented. SoftReMish was evaluated against popular activation functions including ReLU, Tanh, and Mish by replacing the activation function in all trainable layers. The model performance was assessed in terms of minimum training loss and maximum validation accuracy. Results showed that SoftReMish achieved a minimum loss (3.14e-8) and a validation accuracy (99.41%), outperforming all other functions tested. These findings demonstrate that SoftReMish offers better convergence behavior and generalization capability, making it a promising candidate for visual recognition tasks.