My dotfiles and some tests and small learning projects
A visualizer that calculates and displays the Mandelbrot set on the left and a Julia set on the right. You can zoom/unzoom and change the shown Julia set by moving the mouse on the left side. The images are rendered using the GPU thanks to the CUDA library. The CPU version I made a while ago took ~500 ms to calculate an image (using only one thread), this one takes ~ 20 ms (I have an Intel core i7-8550U and a Nvidia MX150).
A small 2D game that makes cars learn to drive around a circuit using genetic algorithm. This image shows the 5th generation on the training circuit.
A renderer that can render 3D models in the Wavefront obj format using raycasting (calculates transparency and reflections). This image shows a hollow cube with 2 mirror walls, a green glass wall and an open side. There are 2 lights: on the top of the spike in the cube and above the camera.
My very first attempt to make machine learning. It is capable of recognizing 28x28 images representing digits with 15% errors (from the MNIST database). I have done a much better neural network since, in its own repository: https://github.com/KubbyDev/OCR_Neural_Network
The program that was used for this project:
A sort of library that can handle simple logic gates. It can generate logic gates based on a truth table, generate a truth table based on the gate, handle user created gates etc... I have done a much better simulator since, in its own repository: https://github.com/KubbyDev/LogicGateSimulator