Skip to content
View Hubizon's full-sized avatar

Highlights

  • Pro

Block or report Hubizon

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don't include any personal information such as legal names or email addresses. Markdown supported. This note will be visible to only you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
Hubizon/README.md

Hi there 👋

LinkedIn: www.linkedin.com/in/hubert-jastrzebski

I'm a student at Jagiellonian University, studying Theoretical Computer Science

  • 🖥️ Passionate about machine learning and AI research
  • 🦾 Skilled in algorithms, data structures, and problem-solving
  • 🥋 Training jiu-jitsu since the age of seven
  • 📖 Fan of Brandon Sanderson’s books
  • 🧶 Enjoy crocheting in my downtime

My projects:

Desktop app for experimenting with simple neural networks. Includes a token system, leaderboard, and an intuitive JavaFX interface. Built with a clean MVVM architecture and a custom neural net engine written from scratch in Java. Uses PostgreSQL to manage users, models, training data, and tokens.

Research project on detecting dependencies between a child’s and mother’s facial expressions during play. Uses PyAFAR to extract facial Action Units for temporal and relational analysis.

Research paper on continual learning, currently available on arXiv and awaiting review. I implemented the core logic, wrote most of the Results section, came up with the name, created the teaser, and designed most of the plots in the paper.

FeNeC is a continual learning method that employs feature clustering and neighborhood-based classification to mitigate catastrophic forgetting in class-incremental learning scenarios. It leverages frozen feature extractors and uses the Mahalanobis distance for classification.

This project serves as a training exercise where I explore various techniques for generating images using methods like autoencoders, GANs, and diffusion models. The goal is to experiment with different approaches and deploy the results on a web interface (click here to try!).

A small C++ library inspired by PyTorch, enabling users to create simple models for deep learning. This project deepened my understanding of gradient descent by implementing it from scratch in C++. It explores the fundamental principles of optimization and backpropagation.

Android game I developed during primary school, deployed on Google Play for a time. The game is a life simulator where players can explore different paths, such as pursuing a career, building relationships, or choosing unconventional lifestyles, all while managing basic needs like food and rest.

This repository contains my algorithmic journey, including problems and solutions from high school to university. It features implementations of algorithms, data structures, and notes for competitive programming and Olympiad preparations.

Popular repositories Loading

  1. cpp-tensor cpp-tensor Public

    C++ 2

  2. coding-archive coding-archive Public

    Jupyter Notebook 1

  3. game-of-life game-of-life Public

    A life simulation Android game where you choose your own path.

    Java

  4. Hubizon Hubizon Public

  5. generative-ai generative-ai Public

    Jupyter Notebook

  6. ai-playground ai-playground Public

    PLpgSQL