A video compression tool made for HackTX 2022

Check out the frontend and the machine learning part

You need to clone ByteSizeML into a folder inside this directory called ByteSizeML. The "results" and "figure" folders should be moved up one level into the ByteSizeBackend folder, as should colorization_deploy_v2.prototxt, colorization_release_v2.caffemodel, and pts_in_hull.npy.

Install the requirements for both ByteSizeBackend and ByteSizeML into the same venv

This includes also installing PyTorch from your own tailored install command provide on the PyTorch Website. Prefereably using Pip. For future reference, this project was created using PyTorch 1.12.1

To run:

flask run