Giters

Simeonedef / Galaxy-Image-Gen

This project proposes a solution to score how 'realistic' sparse galaxy images are as well as a 2-stage GAN to generate those sparse galaxy images.

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Add a high level description of project in README (after finishing report)

AndrejIvanov1 opened this issue · comments

commented

From howto-paper:
" Describe how the results presented in your paper can potentially be reproduced" which i think is not necessary clear

We should mention:

  • put images from kaggle on data/
  • put weights in file pretrained_weights/ → maybe not necessary if we add them there in the submission

I) generators

  • run script to patch images
  • run file to train generator
  • etc

II) regressor

  • run this file for feature extraction
  • run this file for training
  • run this for evaluation

Just to make clear the steps in case they want to try something out

  • mention that for all notebooks the paths and name of the files should be updated accordingly
commented

also add to readme info about :
visualize_clusters.py in analysis/README
and visualize_clusters/ in analysis/README

add readme to common/
add readme to feature_extraction → maybe not necessary as name is self explanatory

commented

These issue kinda became one with #35