Here you can explore my projects and skills!
I work with various programming languages, including C++, Python, and C#. I also use tools such as Docker.
I have experience in low-level programming in assembler and C, which allows me to better understand computer system architecture and write efficient programs.
As a full-stack web developer, I have experience creating applications in production environments, both on the front-end and back-end side.
My passions include artificial intelligence programming and the development of text-based imaging systems. I am also currently learning computer graphics.
I am also interested in IT in general, including server configuration and administration on Ubuntu.
I have experience programming Arduino microcontrollers and creating electronic devices, which allows me to combine programming with electronics.
I have 2 years of professional experience in the programming industry. :)
This project, based on Python and EasyOCR, detects numbers on images and marks them with green boxes. The user can select an image area for analysis or scan the entire image. The program allows editing the results: adding, removing, and modifying selections, and then saving results as a PDF file with a legend and an image with annotations. Additionally, the project can be compiled into a .exe file, making it runnable on computers without Python installed.
"Mystt" is an application for recognizing words based on voice recordings. It uses XGBoost for sound classification, represented by MFCC (Mel-Frequency Cepstral Coefficients) features extracted with librosa. The project includes three main stages: data collection, model training, and result testing.
camera_ai is a project based on an XGBoost classifier, designed to recognize whether a hand is present in real-time camera input. The model is trained on image features, such as grayscale images resized to 64x64 pixels. Users can collect data, label images (e.g., "hand" or "no hand"), and then use the model for real-time predictions. The project allows testing directly via the camera, showing whether a hand is detected on screen.
signAI is a project built with Django framework, enabling recognition of digits in 20x20 pixel images. The application allows users to train the model from scratch, starting from index 0. Once launched locally on the server (127.0.0.1:8000), users can add data and test the model.
Email: mateuszblaszczyk36@gmail.com
Github: https://github.com/Kitler174