Face Classification and Verification

This repository contains code to implement image classification and verification using a ResNet50 network.

Machine Learning Library

The code is written in PyTorch.

Essentials

The code is contained within the tools\ directory. There are five modules, whose functions are as follows:

• test_vrf.py: Takes two images as input, then outputs the cosine similarity between the both.
• test_cls.py: Takes an image as input, then outputs the label.
• runner.py: Trains the network.
• model.py: ResNet50 network.
• dataset.py: Implements the dataset class for both image classification and verification.