/Facial-Similarity-with-Siamese-Networks-in-Pytorch

Implementing Siamese networks with a contrastive loss for similarity learning

Primary LanguageJupyter Notebook

Facial Similarity with Siamese Networks in Pytorch

You can read the accompanying article at https://hackernoon.com/one-shot-learning-with-siamese-networks-in-pytorch-8ddaab10340e

The goal is to teach a siamese network to be able to distinguish pairs of images. This project uses pytorch.

Any dataset can be used. Each class must be in its own folder. This is the same structure that PyTorch's own image folder dataset uses.

Converting pgm files (if you decide to use the AT&T dataset) to png

  1. Install imagemagick
  2. Go to root directory of the images
  3. Run find -name "*pgm" | xargs -I {} convert {} {}.png

Installing the right version of PyTorch

You can go over to http://pytorch.org and download version 0.3.0.post4 for your setup. For Cuda 8 on linux it can be installed using : pip install http://download.pytorch.org/whl/cu80/torch-0.3.0.post4-cp36-cp36m-linux_x86_64.whl

You can find other project requirements in requirements.txt , which you can install using pip install -r requirements.txt

This project requires python3.6