Landscape Colorizing Project (GAN with feature losses)

davilirio · August 24, 2020, 7:04pm

Hey, guys!

Just published this project using fast.ai V1 for landscape coloring using the Landscape Pictures dataset from kaggle! The idea is based on the DeOldify project from lesson 7 2019.
I also created a public album

Here are the links:

github.com

Davilirio/Neural-Nets/blob/master/gan_feature_loss.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "gan_feature_loss.ipynb",
      "provenance": [],
      "collapsed_sections": [],
      "authorship_tag": "ABX9TyPfmWRBHAgPLfjV+dcScEbB",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "markdown",

This file has been truncated. show original

https://drive.google.com/drive/folders/1b4Yy5yvlB2qFeRdFdTswcUlW2BD1DaLH

Any ideas on how to make this project better? All suggestions will help!