Hi, I'm Daniele Bertagnoli
Computer Scientist and Developer
Shortly, I'm a developer who's currently specializing in Machine Learning and Deep Learning topics. My passions are coding, technology and motorbikes! Here you can find my CV, contacts and some of my projects!
Personal Projects
Here, you can find some of the projects I'm currently working on. I'll do my best to keep the list updated; however, you can always drop me an email if you want to know more about them!
Deep Learning Notes
These notes were taken during the years in a self-learning process. You can find general deep learning notions but also more deep and detailed sections dedicated to well-known architectures.
Reinforcement Learning Notes
These notes were taken while I was studying RL by my own. In particular, these notes were taken while reading the book "Reinforcement Learning: An Introduction" published by Richard S. Sutton and Andrew G. Barto. These are basic and also quite outdated notions, however they also represent the basis for the RL approach.
Varjo Communication Framework
This project offers a simple yet efficient way to retrieve raw camera images from a Varjo visor. It is part of a larger GitHub project, IndustrialPoseEstimationFramework. In the linked project, there is functionality to run a pose estimation model in inference mode through a client-server architecture. However, in that setup, the client uses a standard webcam to capture images. With this script, we can integrate raw camera images from the Varjo visor as input for the model.
Linux Dotfiles
This project was developed to simplify and speed up the Linux installation and configuration process. Currently, the dotfiles support two Linux distributions: Ubuntu and Arch-based systems. For Ubuntu, the configuration primarily involves removing unnecessary packages, such as Snap, and applying some styling themes. For Arch-based systems, the configuration installs all the required packages and then applies styling elements. The chosen desktop environment for Arch is Hyprland.
Remote Game Emulator
This project facilitates console emulation on Linux and Windows, featuring a game save system. Users specify a server for file storage, enabling seamless download and upload of game files.
Synthetic Video Generation
This project allows users to generate synthetic videos from CAD models, including .npy files with additional information. Models are loaded dynamically into a Blender scene, and the camera smoothly moves along spherical points to create the final video.
FantasyFootball-UI
This project consists of a dashboard that can be used during fantasy football bids. With this tool, you can keep track of every player sold, as well as the current players' credits and their players.
DragonBlock
The primary objective of this project is to develop a decentralized application (DApp) on the Ethereum blockchain, facilitating a crowdfunding system. Users have the flexibility to either create their fundraising campaigns or contribute to existing ones. Key features include MetaMask integration for secure wallet interactions, the creation of DSTs (Dragon Sphere Tokens) to empower users, the ability to participate in crowdfunding activities, and the implementation of Truffle for streamlined smart contract development and testing.
Vision Odissey
Vision Odyssey is an innovative project that leverages advanced computer vision and machine learning techniques for creating a unique gaming experience. It includes facial and emotion recognition for user registration, login, and dynamic difficulty adjustment, as well as gaze and head pose tracking for character control, offering accessibility to individuals with partial or total paralysis. The game, developed in Unity, communicates with Python-based servers that analyze webcam inputs to control the game and adjust difficulty, utilizing datasets and state-of-the-art algorithms for its functionalities.