2020
Money Grabber AR - ML
Game Programmer

Developed for the course "Sviluppo di Applicazioni per dispositivi mobili" and "Artificial Intelligence for Videogames"  @Università degli Studi di Milano
Made with Unity3D, Unity MLA, Tensorflow. Everything was made in about 2 months.

This project was initially two separate projects: a game in AR and a simulation of an AI using Machine Learning, but since they were complementary i decided to join them and to develop the ML part based on the AR part.

The AR part is an Augmented Reality game in which the player after scanning the environment around him, must collect some money randomly spawned on the ground by physically going on the spawn point. There are 2 different modes: collect as much as you can in 1 minute, or collect 8 coins in the shortest time possible.

The ML part is an advanced AI trained using machine learning that after receiving an environment as input tries to find all the money inside that environment, so this part can be integrated in the AR game.

State of the game
Playable (Android APK Demo, ML part not included) 

Download APK
Github (AR part)   Github (ML part)


Money Grabber AR_.pdf

Documentation (AR part)

Deep RL-trained Agents in Unity.pdf

Documentation (ML part)