Overview / Usage

Classification of Items based on their similarity is one of the major challenges of Machine Learning and Deep Learning problems. But we have seen good results in Deep Learning comparing to ML thanks to Neural Networks, Large Amounts of Data and Computational Power. We have mostly seen that Neural Networks are used for Image Detection and Recognition etc. But I am not going to explain how we execute it. But in brief, we can say that “ Neural Networks are extracting features by finding the patterns in a large set of images using mathematical computation for detecting and recognizing“.

But why Siamese Neural Networks ? and What are Siamese Neural Networks ? . For many of them, it is a new word. Don’t worry! I will make it clear and easy for you. I will also explain the same with a small mini-project. But before that, you should understand some terminologies and topics behind it. Let's dive in!

Methodology / Approach

• Using Siamese Neural Network
• One-Shot Learning
• Contrastive Loss or Lossless Triplet Loss

Technologies Used

* Tensorflow and Tensorflow Lite

* Pytorch

* Python

Repository

https://colab.research.google.com/drive/1FsixLon5Zz3_rFA0xIHzc8Tvnnw8FLr8