Rafael Diniz

A Reinforcement Learning agent that uses Proximal Policy Optimization and Computer Vision to learn how to aim in Counter-Strike

An Image/Video preprocessing library written in CUDA. Currently still extremely early in development. Mostly an independent project that I use to learn CUDA.

Development of a Solar Radiation Map Using Artificial Intelligence Techniques - A Python 3 implementation. The scientific paper for this implementation can be found at https://www.sistema.abcm.org.br/articleFiles/download/36482

Meus arquivos para a matéria de Redes de Computadores 1 (RC1) da PUC Minas com o professor Max do Val Machado.

Catálogo público dos meus códigos feitos em tarefas e atividades das matérias que cursei durante minha graduação em Ciência da Computação na PUC-MG.

Experimental video encoder/decoder built from scratch in PyTorch featuring tensorized block matching, residual coding, scene-aware I-frame selection, chroma subsampling, custom serialization and Zstandard compression