This project’s focus is on in-game Minecraft image conversion. Using self-made paired image data from Minecraft, we attempt to convert between weather conditions, like rainy and sunny, and between different times of day. Using a deep neural network with a U-net architecture, our machine learning model uses paired data to predict and generate images given each pair.

This project uses pix2pix in Pytorch as a reference to build the models.

---

Source code: https://github.com/elaine-ho/Forever_Machine_Learning

Reports:

• Proposal
• Status
• Final

Members:

• Elaine Ho
• Kai Malloy
• Nathan Ma

Group Meeting Time: 7:30PM on Mondays