Professor · Researcher · Free Software Advocate
Building open technologies that transform knowledge into education, research and social impact.
Research Projects
Selected projects in Artificial Intelligence, Deep Learning and Open Source research.
Framework for model extraction attacks against Convolutional Neural Networks using unlabeled natural images. It includes the original method, PyTorch implementation, datasets, trained models and the Copycat CNN Explainer.
Artificial Intelligence ecosystem using LLMs, RAG and open-source technologies to strengthen the regional autism support network.
Research on reducing the amount of labeled data required for practical model extraction attacks.
Interactive visualization of model extraction using CNN Explainer adapted for the Copycat method.
Open Source
Software developed to promote learning, research and knowledge sharing through Free Software.
GNU/Linux server administration platform developed for educational institutions.
Translation and Unicode support for a logic teaching application.
A playful fake AI system created to introduce undergraduate students to Artificial Intelligence concepts.
Interactive material for teaching pointers in C programming.
Teaching recursion through the implementation of Minesweeper.
Educational material introducing Artificial Intelligence and Large Language Models.
Research
My research focuses on Artificial Intelligence, Deep Learning and Open Source technologies, especially techniques that make intelligent systems more accessible, understandable and useful for society.
Open-source LLMs, RAG and local AI.
CNNs, image classification and explainability.
Model extraction and knowledge transfer.
GNU/Linux and Open Source ecosystems.
Intelligent systems applied to education and society.
AI supporting inclusion and accessibility.
Journey
More than twenty years learning, building and sharing knowledge through Free Software, Artificial Intelligence and education.
Started programming and discovered that software is much more than technology: it is a way to create knowledge.
Joined the Free Software community and began contributing to open technologies.
Contributed to the development of the Muriqui Linux distribution adopted by the Brazilian Ministry of Education (MEC).
Worked as developer and later team manager of Vix Linux, deployed in public schools in Vitória, also training teachers to use Free Software in education.
Started teaching Computer Science while conducting research, supervising students and developing open educational software.
Completed the Ph.D. with the Copycat CNN framework, releasing datasets, software and educational tools as open source.
Developing local Large Language Model infrastructures and intelligent systems to strengthen autism support through Free Software.
About
I am a Professor at the Federal University of Espírito Santo (UFES), where I work on Artificial Intelligence, Open Source Software and Computing Education.
My work combines research, software development and extension activities to transform knowledge into technologies that benefit society.
I believe Free Software is much more than open source code. It is a way to create, share and democratize knowledge.
Publications
Research contributions in Artificial Intelligence, Deep Learning, Computer Vision and Model Extraction.
Pattern Recognition
Extended journal version of the original Copycat CNN method, demonstrating that unlabeled natural images can successfully perform model extraction attacks.
IJCNN
Original proposal of the Copycat CNN method presented at the International Joint Conference on Neural Networks.
Ph.D. Thesis
Doctoral thesis describing the complete framework, datasets, experiments and software developed during the research.