There is a newer version of the record available.

Published April 2, 2025 | Version 2.2.0

CNN Parameter Tuner 2.2.0

  • 1. ROR icon Czech University of Life Sciences Prague
  • 2. ROR icon University College London

Description

Greetings and welcome to CNN Parameter Tuner 2.2.0!

This innovative application is your gateway to the fascinating world of Convolutional Neural Network (CNN) parameter tuning. Unlock the full potential of image analysis by fine-tuning CNN algorithms with precision and ease.

Enjoy a wide range of preprocessing options, including normalization, standardization, and augmentations, as well as the flexibility to experiment with various train-test ratios, custom CNN models, epochs, batch sizes, activation functions, optimizer functions, and loss functions.

Please ensure that the images are organized into separate folders, with the folder names representing the class labels for the classification task. Once the data is loaded, the software will automatically recognize the folder names as the class labels.

Kindly note that the software may take approximately one minute to appear. We appreciate your patience!

Update: We are pleased to introduce various preprocessing options, including normalization, standardization, and augmentations, to further enhance your experience!

 

Files

cnn tuner book v2.2.0.pdf

Files (734.9 MB)

Name Size
md5:7252d4925f275c94a2c2c30b8f52833b
728.9 MB Download
md5:608fd06f256857172e25bb7d49faf810
6.0 MB Preview Download

Additional details

Related works

Has part
Journal article: 10.1016/j.ecoinf.2024.102932 (DOI)

Software

Repository URL
https://github.com/Gokultcr/CNN-Parameter-Tuner-1.0
Programming language
Python
Development Status
Active