Vinicius Arruda
Bio

I am a Machine Learning Engineer working across computer vision, NLP, and LLMs. My academic research was carried out at the Federal University of Espírito Santo in the High Performance Computing Laboratory, advised by Prof. Thiago Oliveira-Santos. It focused on unsupervised domain adaptation for object detection using synthetic target images, image-to-image translation, and neural style transfer.

I am currently interested in representation learning, world models, vision-language models, and generalization. I am actively looking for PhD opportunities, research collaborations, visiting student positions, research internships, and applied scientist roles.

Publications
Cross-Domain Object Detection Using Unsupervised Image Translation
Expert Systems with Applications (ESWA), 2022
Vinicius F. Arruda, Rodrigo F. Berriel, Thiago M. Paixão, Claudine Badue, Alberto F. De Souza, Nicu Sebe, Thiago Oliveira-Santos
[Paper]
Cross-Domain Car Detection Using Unsupervised Image-to-Image Translation: From Day to Night
International Joint Conference on Neural Networks (IJCNN), 2019
Vinicius F. Arruda, Thiago M. Paixão, Rodrigo F. Berriel, Alberto F. De Souza, Claudine Badue, Nicu Sebe, Thiago Oliveira-Santos
[Paper] [Project Page] [Code] [Demo Video]
Particle Swarm Optimization for training artificial neural networks of type ELM: A case study for time series prediction
XLVIII Brazilian Symposium of Operational Research, 2016
Vinicius F. Arruda, Renato A. Krohling
[Paper]
Posts
A Brief Introduction to Artificial Neural Networks
Jan 30, 2019
An introduction with numerical examples solving logical gates and a simple training with a genetic algorithm to play Flappy Bird.
A Brief Introduction to Genetic Algorithm
Sep 19, 2017
Inspired by Darwin's theory of evolution, the genetic algorithm has the target problem modeled as a chromosome that evolves over time.
A Brief Introduction to Cellular Automata
Aug 20, 2017
A cellular automaton is a collection of cells on a grid that evolves through discrete time steps according to neighboring cell states.
"What I cannot create, I do not understand."
— Richard Feynman
adapted from genforce.github.io
inspired by vsitzmann.github.io