Alessandro Sestini

DeepCrawl. Deep reinforcement learning for the development of RogueLike games.

Parallel-K-Means. A parallel implementation of K-Means algorithm in C++ and OpenMP.

Policy-Fusion-RL. Policy fusion methods to combined pre-trained policies without the need of retraining them, mainly for game developers. Unlike systems built to replace human players, our agents aim to produce meaningful interactions with the player, and at the same time demonstrate behavioral traits as desired by game designers.

CCPT. A deep reinforcement learning algorithm to perform automatic analysis and detection of gameplay issues in complex 3D navigation environments. CCPT method combines curiosity and imitation learning to train agents to methodically explore in the proximity of known trajectories.

Facial-Expression-Prediction. A simple python program that can predict different facial expressions given a neutral model.

Adaptive-NPCs-with-procedural-entities. Deep policy networks for NPC behaviors that adapt to changing design parameters in Roguelike games.

Parallel-Histogram-Equalization. A simple parallel histogram equalization in CUDA and Java.

Conway-s-Game-of-Life-Python. A Conway's Game of Life implementation in Python.

GeyserCG-3D. Design and simulation of geyser explosions through Computer Graphics.

rl_sesto. My implementation of Proximal Policy Implementation

SeeForMe. A mobile App for the automatic recognition of museum artworks and the semiautomatic management of multimedia feedback.

Wesnoth-Companion-App. An Android application based on the game “The Battle for Wesnoth” for Human Computer Interaction project.

Worldmodel-Visualizer. A 3D visualizer to understand training and behavior of an exploratory agent.

Demonstration-Efficient-AIRL. A technique based on Adversarial Inverse Reinforcement Learning (AIRL) which can significantly decrease the need for expert demonstrations in PCG games. Through the use of an environment with a limited set of initial seed levels, DE-AIRL is demonstration-efficient and able to extrapolate reward functions which generalize to the fully PCG domain.

Conway-s-Game-of-Life-Cpp. A Conway's Game of Life implementation in C++.

VisCCPT. Official codebase for VisCCPT Visualization Tool.

Navigation-Environment. Playtesting Env made by Ale and SEED.

Godot-RL-Env. I've been working many years with prototype environments in Unity, but I've never worked with Godot. This is my attempt to create a prototype environment in Godot for RL training.