Skip to content

akshaysharma096/Siamese-Networks

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

6 Commits
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Siamese Neural Networks for Few Shot Learning

The problem has been inspired by Fellowship.AI, demo challenge

  • The model has been trained using tensforflow backend in Keras.

  • It tries to solve the problem of image verification when the quantity of data available for training Deep Learning models is less.

  • The model has been implemented to solve the problem based on the paper by Gregory et. al Siamese Neural Networks for One-Shot Image Recognition.

    • The ideas from the paper have been used to extend the model for few shot learning.
  • Omniglot, dataset has been used for training the model

    • The dataset has 1623 character classes, each with 20 examples.
    • The model tries to build a few shot classifier using the ides presented in the paper by Gregory et. al for the dataset.

Reading