Sabrina Alvarenga Email and Phone Number

Studied Computer Science at the Federal University of Minas Gerais, currently enrolled in a Master's Degree at UFMG.In the academy, I'm currently working on graph theory and computational complexity, with a focus on reconfiguration problems in graphs, and focusing on scientific efforts. I am also active in the industry, where I currently contribute to software and mathematical modelling of solutions regarding nanotechnology, for instance, nanospectroscopy, which is used for studying materials such as graphene, at FabNS.My current interests involve pure and applied mathematics, such as topology, algebra, combinatorics and their intersections, and also the scientific applications of those techniques in fields such as experimental physics and data analysis.Some of my past experience:- Teaching assistant in Discrete Mathematics at UFMG (2022)Provided support to Professor Jeroen van de Graaf in a discrete mathematics course for first and second-year students. My role involved tasks like correcting, grading, and assisting with problemsets—both online and in person. Additionally, had the opportunity to share insights and teach classes on topics like enumerative combinatorics, probability and graph theory throughout the course.- Relighting human avatars for videos/motion transfer (2019-2020)The motion transfer project is organized by the Computer Vision and Robotics (VeRLab) at the Federal University of Minas Gerais and it aims to reproduce a human movement with a model different than the original, preserving body shape and size constraints. The relighting is meant to improve the motion transfer’s realism, by adding a new shading to thehuman. This involves inverse rendering to compute the human geometry and the light applied to it and rendering and light modelling techniques. This is based of spherical harmonics, since the bands’ coefficients can be learned by a reverse rendering neural network.- Multiview project (2020)This project is being developed as a collaboration between VeRLab and Petrobras and its goal is to build an auxiliary framework for oil rig construction and maintenance. The techniques used are 3D reconstruction based on scans by various sensors, and point cloud rendering. Techniques used for dealing with point clouds include geometric alignment with CAD models through transform matrices, and shading techniques for better visualization of the point clouds using Three.js.