Final Projects Spring 2021 - Applied Machine Learning for Games


AgentSmart - Autonomous Driving Agent
Website is here.
Google Drive is
here.

Bomb in hand - is a combat game where the player controls a character through gestures and fight against bots across a deformable landscape. The player must survive and throw bombs to destroy the opponent‘s assets.
Website is here.
Google Drive is
here.

Crossy Road - our team collaborated to train an intelligent chicken to cross as many as roads as she can in the game, and win the highest score.
Website is here.
Google Drive is
here.

Dark Forest - is an open-source 3D interactive battleground and survival sandbox game.
Website is
here.
Google Drive is
here.

Doodle Jump - guide a four-legged creature called "The Doodler" up a never-ending series of platforms without falling.
Website is
here.
Google Drive is
here.

Free Run - is a single-player, 3D, third-person perspective game. One plays this game as a “Raccoon” that overcomes the obstacles that come in its way.
Website is here.
Google Drive is
here.

LingoRobo - enabling AI agents following natural language instructions.
Website is here.
Google Drive is
here.

Obstacle Tower Challenge - is a high-fidelity game that operates in 3D, generates new visuals and themed layouts with each level and game using “Procedural Content Generation”, has a high branching factor, and requires agents to learn to solve sub-tasks.
Website is here.
Google Drive is
here.

Park It Right! - The focus of this project is to develop an agent that is adept in obstacle avoidance and seamless navigation.
Website is here.
Google Drive is
here.

Phoenix - an open-source, reproducible and interpretable Mahjong agent.
Website is here.
Google Drive is
here.

qDeerein - Here we leverage Deep Q learning and Actor-Critic to train agents and make them play on their own.
Website is here.
Google Drive is
here.

The Snake Game - we re-create the classic Snake Game by changing the way it should be played. Gamers can now use hand gestures as well voice commands to play this game, adding a completely new dimension to the gaming experience and the fun!
Website is here.
Google Drive is
here.

Text Room Adventure - we incorporate pre-trained language models into the deep Q-learning framework to bring in rich commonsense knowledge existing in external text corpus, and combine history actions into decision making process.
Website is here.
Google Drive is
here.

ThinkBot - a physics-based driving game.
Website is here.
Google Drive is
here.

Trojan Knights - LiteChess is a low-resource, neural network based chess engine written in Python that plays at the 2000 ELO level. The neural network architecture is inspired by the DeepChess algorithm and the engine backend is adopted from the popular Sunfish engine.
Website is here.
Google Drive is
here.

Virtual Equipment System - exploring VR GUIs with machine learning.
Website is here.
Google Drive is
here.

VizDoom Bot - rhe ViZDoom platform makes use of an engine based on the video game Doom to provide a physics-based 3D
environment similar to real world environments for machine learning.

Website is here.
Google Drive is
here.

TOP