I'm Paolo Bonomi, a computer scientist from Milan, Italy. I’m interested in computational geometry, machine learning, and software engineering—especially where these fields connect.
I really enjoy working with complex data and building new machine learning models. That interest led me to the world of Computer Vision, where I like solving problems that involve images and patterns.
I'm a curious person by nature, and I always want to understand how things work under the hood—not just use libraries or tools, but really learn what’s happening behind the scenes.
Some of the topics I like to explore:
- Deep Learning (like CNNs, Transformers, Reinforcement Learning)
- Computational Geometry
- GPU Programming and Making Code Run Faste
- Indie Game Development
- Math and Physics
CapsNetwork – A PyTorch implementation of Capsule Networks, based on Hinton's dynamic routing architecture, designed to improve spatial understanding over traditional CNNs. Includes tools for training, evaluation, and result visualization on MNIST.
DroneDelivery – A Java-based prototype for a distributed drone delivery system. Utilizes REST APIs and MQTT messaging via the Mosquitto broker, with Gradle for build automation. Simulates real-time drone task coordination.
Coons Patch Visualizer – An interactive web app built with JavaScript and p5.js to visualize Coons patches. Allows manipulation of boundary curves and blending functions for educational exploration in computational geometry.
vPhysic – An C++ project that implements a basic Verlet physics engine using OpenGL for real-time visualization. It simulates particle systems and physics constraints in an interactive graphical environment.
ExploRace – An online multiplayer game prototype developed in Unity using C#. It focuses on exploration and racing mechanics, showcasing basic networking and gameplay features.