David's Projects
Applying chroma key to images/videos, regardless of the colour.
Let us control diffusion models
Testing a small API REST with Behave, Selenium and Allure.
A simple API REST made with Flask in order to generate custom passwords. Hosted at PythonAnywhere.
Convolutional Neural Networks
A python package for embedding pandas DataFrames as images into pdf and markdown documents
Jupyter notebooks for the code samples of the book "Deep Learning with Python"
Official Code for DragGAN (SIGGRAPH 2023)
A web scraper to do automated massive image downloads using Selenium. Tested on Firefox, but should work on Chrome/Edge with their respective webdriver.
Homography operations applied to insert images inside others regardless of the perspective.
Jupyter Notebook to store and evaluate long-run evolution of house sales, based on Idealista website.
A web scraping algorithm to get key data from house sales in a few seconds. Easy to modify in order to check any location.
A simple Ragnarok Online MVP respawn tracker. It tells when will it come back, even if its not time-based, but based in specific number of mobs killed (Ex: Shiny Teddy Bear).
An algorithm combining automation and artificial vision to find shiny pokemon inside this fangame. It's pretty much an example of how the mix of those kind of technologies can automate many stuff.
A collective list of free APIs
Automated tests and reports performed on "https://highlifeshop.com/speedbird-cafe" using Selenium, Behave and Allure.
A simple website to keep track of MVP kills on Ragnarok Online, made with Django. Includes user authentication to save and retrieve individual data from PostgresSQL DB.
Classifying traffic lights on a video streaming with OpenCV, classic computer vision using HSV channels to determine colors.
Repairing old photos with OpenCV, starting from the masking process all the way up to the inpainting algorithm from Telea. This can also be used to just remove undesired items from images.
Hello there!
Coding and testing a convolutional neural network for classifying traffic lights, with Keras and Tensorflow, using LISA dataset.
Traffic lights tracking and color detection with OpenCV. Combination of a MOSSE tracker and inner freehand rectangles.
Train a state-of-the-art yolov3 object detector from scratch!