DESI Legacy Imaging Surveys galaxy classification
Introduction
This project aim to classify the galaxy images collected using DESI Legacy Imaging Surveys using convolutional neural network. We used a ResNet-50 pretrained model which was trained on ImageNet dataset. The model is finetune for 50 epochs and achived 85% accuracy on test dataset.
Installation
Requirements
We have trained and tested our models on Ubuntu 18.0, CUDA 11.0, Python 3.8.
pip install torch==1.7.1+cu110 torchvision==0.8.2+cu110 torchaudio==0.7.2 -f https://download.pytorch.org/whl/torch_stable.html
pip install -r requirements.txt
Dataset preparation
Please download the dataset from here by running follwing command and put in dataset folder.
wget https://astro.utoronto.ca/~hleung/shared/Galaxy10/Galaxy10_DECals.h5