You can find a video of the winning solution [here](https://www.youtube.com/watch?v=OlMG_8NqrdU).

This semester you will get a chance to try your hand at detecting fake images from real images.  I trained a [model](https://github.com/jeffheaton/pretrained-gan-minecraft) on images that I collected from the [Minecraft video game](https://en.wikipedia.org/wiki/Minecraft).  From the provided link, you have access to my trained model, and can generate more fake data, if you like.  However, if you would like additional real data, you will need to capture it from Minecraft yourself.

The following is a real image from Minecraft:
![Real Minecraft](https://github.com/jeffheaton/jheaton_images/blob/main/kaggle/spring-2021/mc-34.jpg?raw=true)

This Minecraft image is obviously fake:

![alt](https://github.com/jeffheaton/jheaton_images/blob/main/kaggle/spring-2021/mc-202.jpg?raw=true)

Some images are not as easily guessed, such as this fake image:

![alt](https://github.com/jeffheaton/jheaton_images/blob/main/kaggle/spring-2021/mc-493.jpg?raw=true)

You will also have to contend with multiple times of the day.  Darker images will be more difficult for your model.

![alt](https://github.com/jeffheaton/jheaton_images/blob/main/kaggle/spring-2021/mc-477.jpg?raw=true)