Silvério Pereira

SILVÉRIO PEREIRA R- ‡ github.com/silveryu ¯ linkedin.com/in/silveriopereira ABOUT A Computer Engineer MSc with interest in Machine Learning and surrounding technologies. With experience in developing software for Deep Learning, Distributed Systems and Data Science in an academic environment, through which i acquired skills in: • Machine learning: Keras, Tensorflow, Scikit-learn • DevOps: Linux, Git, Bash, Docker • Data Wrangling: MySQL, PostgreSQL, Kafka • HTTP APIs: Django, Flask • Programming languages: Python, Java, C, C++, R EDUCATION Universidade de Aveiro Sep 2014 - Jul 2017 Bsc in Computer and Telematics Engineering. Which has a focus on Security, Computer Architecture, Computer Networks and Artificial Intelligence as well as other Computer Science topics. Universidade de Aveiro Sep 2017 - Feb 2021 Msc in Computer and Telematics Engineering. Where i focused mainly in Machine Learning, Security and Distributed Systems. My Msc Thesis focuses on Anomaly Detection, Clustering methods and Approximate Nearest Neighbours. PROJECTS Audio and video codecs Some audio and video codecs that use various types of compression algorithms developed in C++ High level content-based image retrieval A content-based image retrieval system using solely high-level features. It makes use of Object Detection, One-shot Face Recognition, Visual Saliency and Optical Character Recognition. VHDL Neural Net A Feedforward neural network with custom hardware accelaration Checkers A checkers modeling program able to use a live, non static video feed. Made using only traditional computer vision techniques. Pokedex A pokedex android app, developed for a Human-Computer Interaction class. Ray tracing A simple ray tracing tutorial and animation developed in WebGL. Ekains An A*-based snake playing AI agent Approximate Growing Neural Gas An alteration of the Growing Neural Gas algorithm, a clustering algorithm in the same vein of K-means. Commonly used for 3D reconstruction and compression purposes. It was altered to use Approximate Nearest Neighbours techniques, making it useful in high dimensional settings. Main subject of my master’s thesis. HACKATHONS Some other quick projects and competitions i’ve been involved in: Markovitter Generational model based on a trained markov chain that outputs quotes for any twitter account. Done at Hacktivate in Feb 2018 (1st place) Dhoraryus Schedule helper for university classes using constraint search. Done at Make or Break in Sep 2017 Battle of Universities National University team challenge done by Outsystems. Consists of solving several computing and logic problems. Done in Nov 2018 (2nd Place)