Summary of the Project

Our project aims at converting screenshots of the view of character in Minecraft into the pictures in real world. The screenshot is based on the random game scene that has been settled. The input is the screenshot of the scene in Minecraft, a world which everything is made of cubes. And what we need to do is to use deep learning techniques to transform the screenshot to a picture composed by items and scenes similar to those in real world. In the generated picture, we smooth the vertically unreal frames of items and match the items in Minecraft with its real world symbolization. By doing this we can make game player enjoy playing game in a more realistic scene that can bring a whole new game experience.

AI/ML Algorithms

For our project, we plan to use generative adversarial network (GAN) to build our machine learning system.

Evaluation Plan

To determine the success of our project, we will evaluate our machine learning system both quantitatively and qualitatively:

Appointment with the Instructor

9:30am on Monday, April 29 (DBH 4204)