Alex J. Champandard

Super Resolution for images using deep learning.

A Python research tool from 2016 that uses a neural network to upscale low-resolution photos by 2x or 4x, or to remove JPEG artifacts and noise, the network guesses plausible detail rather than recovering what was originally there.

Turn your two-bit doodles into fine artworks with deep neural networks, generate seamless textures from photos, transfer style from one image to another, perform example-based upscaling, but wait... there's more! (An implementation of Semantic Style Transfer.)

Render a ShaderToy script directly to a video file.

Opt-Out tool to check Copyright reservations in a way that even machines can understand.

Generate unique 2D textures from example images and a specification with this command-line tool.

No description.

NanoGPT (124M) in 5 minutes

Experiments in Python applying deep learning to PlanetWars.

🚃 Automated task/test framework for writing differentiable software.

Make a detailed summary of data used during AI/ML training.

Cython bindings and Pillow plugin for JPEG XL

No description.

No description.

My experiments with AI for THE RESISTANCE.

No description.

An opinion repository about the state of generative machine learning.

Easily turn large sets of image urls to an image dataset. Can download, resize and package 100M urls in 20h on one machine.

Website and blog for creative.ai.

My development branch for VisPy

An open source implementation of CLIP.

Adding JXL support.

No description.

No description.

No description.