Hello internet. I’m a Ph. D. student in Artificial Intelligence. I’m also a computer programmer/software developer.
Currently, my research interests revolve around accelerating deep reinforcement learning (DRL) through high-level, abstract state representations guided by some kind of formalism, such as object-oriented MDPs and qualitative spatial reasoning. I am particularly interested in the applications in real-time strategy games, such as StarCraft 2.
In my research career, I’ve also worked with robots and computerized adaptive tests. In my professional career, I’ve worked with back-end web development and database and server maintenance.
Use the table of contents below or the icons in the footer to find out more about what I am up to.
- Object detection and robotics
- Computerized adaptive testing
- Miscellaneous Projects
Jonas Henrique Renolfi de Oliveira, Isaac Jesus da Silva, Thiago Pedro Donadon Homem, Douglas De Rizzo Meneghetti, Danilo Hernani Perico, Bianchi, and C. Reinaldo A. 2019. Object detection under constrained hardware scenarios: a comparative study of reduced convolutional network architectures. In 2019 XVI Latin American Robotics Symposium and VII Brazilian Robotics Symposium (LARS/SBR) (2019 XVI Latin American Robotics Symposium and VII Brazilian Robotics Symposium (LARS/SBR)).
Ocimar Munhoz Alavarse, Érica Maria de Toledo Catalani, Douglas De Rizzo Meneghetti, and Rodrigo Travitzki. 2018. Teste Adaptativo Informatizado Como Recurso Tecnológico para Alfabetização Inicial. Revista Iberoamericana de Sistemas, Cibernética e Informática 15, 3 (2018), 68–78.
Douglas De Rizzo Meneghetti and Plinio Thomaz Aquino Junior. 2017. Application and Simulation of Computerized Adaptive Tests Through the Package Catsim. (July 2017). Retrieved from http://arxiv.org/abs/1707.03012
Douglas De Rizzo Meneghetti and Plinio Thomaz Aquino Junior. 2014. Técnicas de Clustering Aplicadas a Testes Adaptativos Informatizados.
André de Souza Mendes, Douglas De Rizzo Meneghetti, Marko Ackermann, and Agenor de Toledo Fleury. 2016. Vehicle Dynamics-Lateral: Open Source Simulation Package for MATLAB. In SAE Technical Paper Series (SAE Technical Paper Series). DOI:https://doi.org/10.4271/2016-36-0115
Rodrigo Travitzki, Douglas De Rizzo Meneghetti, Ocimar Munhoz Alavarse, and Érica de Toledo Catalani. 2018. How to build a Computerized Adaptive Test with free software and pedagogical relevance? In International Academic Conference on Teaching, Learning and E-learning (International Academic Conference on Teaching, Learning and E-learning), 117.
Object detection and robotics
During my stay at FEI University Center, I was responsible for integrating object detection techniques into a domestic robot. I worked with libraries such as OpenCV and the TensorFlow Object Detection API as well as ROS, the Robot Operating System. This section provides a centralized collection of resources I’ve created during my years working with computer vision and object detection.
Since you’re here, take a look at the set of Python scripts I provide to help in the creation of TFRecord files and label maps for the TensorFlow Object Detection API, as well as TXT annotation files for YOLOv3. I also authored a brief tutorial on how to train a detection model using the TensorFlow Object Detection API with the help of my scripts.
dodo detector (Python)
dodo detector is a Python package that encapsulates OpenCV object detection via keypoint detection and feature matching, as well as the TensorFlow Object Detection API in a single place. See a simple tutorial here.
dodo detector ros (ROS)
I am also the creator of dodo detector ros, a ROS package that allows dodo detector to interface with USB cameras as well as both Kinects v1 and v2, in order to detect objects and place them in a 3D environment using the Kinect point cloud.
Computerized adaptive testing
For years, I have been a member of the Study and Research Group in Educational Assessment (a free translation of Grupo de Estudos e Pesquisas em Avaliação Educacional, Gepave), where I specialized in the study of Item Response Theory and computerized adaptive tests. While I am not active as of late, I had the opportunity to be a part in a few projects, listed below.
jCAT is a Java EE web application whose purpose is to apply both an electronic version and a computerized adaptive test of Provinha Brasil, a nation-wide educational evaluation for Brazilian students in the second year of basic school.
A Python package that simulates a set of examinees taking a computerized adaptive test. There are different options for initialization, selection and proficiency estimation methods as well as stopping criteria for the test. Useful for studying item exposure and can also be used to power other applications. Documentation here. ArXiv paper here.
FEI LaTeX class
a LaTeX class used by FEI University Center students to author their term papers, masters dissertations and doctoral theses under the institution typographical rules. The document class, a
.tex template file and the PDF documentation are available on CTAN. There is also a wiki for miscellanoues tips and a Google group (in Portuguese) for me to communicate with users.
I have implemented a few machine learning and reinforcement learning algorithms, as well as some interesting numerical analysis procedures and a full-fledged matrix class, all in C++. They may not be directly useful in third-party projects, as they are not as optimized as their commercial counterparts, but they are nonetheless well documented for those interested in learning.