Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Pydimarry, Sai Abhinav, Khairnar, Shekhar Madhav, Palacios, Sofia Garces, Sankaranarayanan, Ganesh, Hoagland, Darian, Nepomnayshy, Dmitry, Nguyen, Huu Phong
Format: Preprint
Veröffentlicht: 2024
Schlagworte:
Online-Zugang:https://arxiv.org/abs/2410.06879
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866916429324877824
author Pydimarry, Sai Abhinav
Khairnar, Shekhar Madhav
Palacios, Sofia Garces
Sankaranarayanan, Ganesh
Hoagland, Darian
Nepomnayshy, Dmitry
Nguyen, Huu Phong
author_facet Pydimarry, Sai Abhinav
Khairnar, Shekhar Madhav
Palacios, Sofia Garces
Sankaranarayanan, Ganesh
Hoagland, Darian
Nepomnayshy, Dmitry
Nguyen, Huu Phong
contents In the field of pattern recognition, achieving high accuracy is essential. While training a model to recognize different complex images, it is vital to fine-tune the model to achieve the highest accuracy possible. One strategy for fine-tuning a model involves changing its activation function. Most pre-trained models use ReLU as their default activation function, but switching to a different activation function like Hard-Swish could be beneficial. This study evaluates the performance of models using ReLU, Swish and Hard-Swish activation functions across diverse image datasets. Our results show a 2.06% increase in accuracy for models on the CIFAR-10 dataset and a 0.30% increase in accuracy for models on the ATLAS dataset. Modifying the activation functions in architecture of pre-trained models lead to improved overall accuracy.
format Preprint
id arxiv_https___arxiv_org_abs_2410_06879
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Evaluating Model Performance with Hard-Swish Activation Function Adjustments
Pydimarry, Sai Abhinav
Khairnar, Shekhar Madhav
Palacios, Sofia Garces
Sankaranarayanan, Ganesh
Hoagland, Darian
Nepomnayshy, Dmitry
Nguyen, Huu Phong
Computer Vision and Pattern Recognition
In the field of pattern recognition, achieving high accuracy is essential. While training a model to recognize different complex images, it is vital to fine-tune the model to achieve the highest accuracy possible. One strategy for fine-tuning a model involves changing its activation function. Most pre-trained models use ReLU as their default activation function, but switching to a different activation function like Hard-Swish could be beneficial. This study evaluates the performance of models using ReLU, Swish and Hard-Swish activation functions across diverse image datasets. Our results show a 2.06% increase in accuracy for models on the CIFAR-10 dataset and a 0.30% increase in accuracy for models on the ATLAS dataset. Modifying the activation functions in architecture of pre-trained models lead to improved overall accuracy.
title Evaluating Model Performance with Hard-Swish Activation Function Adjustments
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2410.06879