Filipe Gonçalves
HEAD OF ENGINEERING, LEAD EXPERT, SOLUTIONS ARCHITECT, DEEP LEARNING SPECIALIST, ASSISTANT PROFESSOR, DATA SCIENCE RESEARCHER
I am a software engineer and solutions architect with expertise in AI development and deployment, focusing on implementing innovative technologies that address real-world challenges. I'm passionate about turning ideas into practical solutions and enjoy leading and mentoring technical teams.
Outside of work, my hobbies include developing personal tech projects, reading books and technical articles, improving physical fitness through daily exercise, engaging in agriculture, socializing with friends and family and travelling. These activities fuel my creativity and enrich my approach to problem-solving.
MAIN DOMAINS
Automated Driving Perception
Automated Driving Perception uses Deep Learning to analyze LiDAR, Radar and Camera data, enabling accurate detection, tracking and recognition of objects surrounding the autonomous vehicles.
Automated In-Vehicle Sensing
Automated In-Vehicle Sensing applies Deep Learning to analyze audio and video data with a focus on passenger safety, enabling real-time recognition of emotions, pose estimation, activity detection, violence identification and weapon detection.
Biometric & Behavioral Analysis
Biometric & Behavioural Analysis uses Machine Learning to non-invasively monitor and interpret user interactions with devices like keyboards and mice. This analysis aims to understand the user's state and adapt system responses to better meet individual needs.
ABOUT FILIPE
SOLUTIONS ARCHITECT & AI SPECIALIST
Kickstarted pre-development for Level 4 Automated Driving Perception in Portugal and currently is the Technical Lead for pre-development of Level 2-3+ Automated Driving Perception. Developed AI-based algorithms for autonomous vehicle perception, specializing in temporal 3D object detection and tracking using Lidar point clouds and camera images.
Advanced in-vehicle sensing and emotion recognition algorithms for ride-sharing and autonomous vehicles. Developed audio-based AI for detecting occupants' emotions and activities and optimized image-based AI for real-time pose estimation and activity recognition on edge devices. Implemented fusion models combining audio and image data to predict occupant activities and anticipate violent acts.
Created AI algorithms to predict students' perceived stress by analyzing mouse dynamics and decision-making during online exams.
Taught AI courses to over 190 MSc students in Computer Science and Biomedical Engineering and software courses to more than 40 BSc students in Physics Engineering at the University of Minho.
Supervised 12 MSc dissertations focused on AI and Deep Learning.
EDUCATION
PhD in Computer Science
School of Engineering, University of Minho
Dissertation: "Human Well-being Monitoring: Intelligent Techniques Application for Improved Accuracy and Privacy"
MSc in Computer Science and Engineering
School of Engineering, University of Minho
Thesis: "Computer-Interpretable Guidelines in Decision Support Systems"
