Skip to content

A custom Convolutional Neural Network (CNN) from scratch for landmark classification and a second CNN using Transfer Learning to benchmark various pre-trained models.

Notifications You must be signed in to change notification settings

DyuthiVivek/Landmark_Classification_and_Tagging

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

4 Commits
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Landmark_Classification_and_Tagging

Created a custom Convolutional Neural Network (CNN) from scratch for landmark classification

Started with an initial Conv2d layer, progressively increasing filters while maintaining 3x3 kernels. ReLU activations and BatchNorm layers were added for non-linearity and normalization. MaxPooling layers were used to downsample the feature maps. Dropout was used to prevent overfitting.

Implemented a second CNN using Transfer Learning to benchmark various pre-trained models

Chose ResNet-18 due to its proven performance and efficiency in various image classification tasks. Since ResNet18 was trained for the ImageNet task, it is a good choice for landmark classification tasks.

About

A custom Convolutional Neural Network (CNN) from scratch for landmark classification and a second CNN using Transfer Learning to benchmark various pre-trained models.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published