Ricardo Cruz

 ricardo.pdm.cruz@gmail.com  rpmcruz  0000-0002-5189-6228  PDF

Ricardo Cruz has worked on a wide range of machine learning topics, with particular emphasis on theoretical aspects of deep learning and computer vision – with 20+ publications and 100+ citations in such topics as: • adapting ranking models for class imbalance; • making convolutional neural networks invariant to background; • making them faster by adjusting the computational effort to each image; • losses for ordinal regression. He is a Post-doc Researcher on autonomous driving at the Faculty of Engineering, University of Porto, and he has been a researcher at INESC TEC since 2015, where his research earned him the computer science PhD in 2021. He has a BSc in computer science and a MSc in applied mathematics. He has frequently been an invited professor at the Faculty of Engineering, University of Porto, where he earned a pedagogic award.

Education

2021
PhD in Computer Science
Joint degree University of Porto, Minho and Aveiro
2015
M.Sc. in Mathematical Engineering
Faculty of Sciences, University of Porto
2012
B.Sc. in Computer Science
Faculty of Sciences, University of Porto

Work Experience

2023/07–present
Post-doctoral Researcher on Autonomous Driving INESC TEC [in partnership with Bosch]
2021/09–2023/06
Post-doctoral Researcher on Autonomous Driving University of Porto (FEUP) [in partnership with Bosch]
Collaboration between the University of Porto and Bosch Car Multimedia to improve autonomous driving perception. I developed frameworks for object detection using camera and LiDAR (2D discretization and raw point-clouds), published new methods for efficient semantic segmentation and ordinal regression, supervised six master's theses, four bachelor's projects, and other team members, responsible for the HPC infrastructure (using Slurm).
2023/09–2024/02
Invited Auxiliary Professor University of Porto (FEUP)
Courses on Machine Learning (theoretical lessons): - OAT4001 , FACVC
2021/09–2022/08
Invited Auxiliary Professor University of Porto (FEUP)
Courses: Programming Fundamentals (practical lessons in Python): - L.EIC003 , Data Structures and Algorithms (practical lessons in C/C++): L.EEC009
2018/09–2021/08
Invited Teacher Assistant University of Porto (FEUP)
Courses: Programming Fundamentals (practical lessons in Python): L.EIC003, Programming (practical lessons in C/C++): L.EIC009
2015/09–2021/08
Research Assistant on Machine Learning and Computer Vision INESC TEC
Research focus: re-thinking fundamentals about image classification and semantic segmentation (8+ publications), in particular (1) a method for background invariance using adversarial training, (2) new losses that minimize absolute trade-offs between Type 1 and 2 errors instead of relative trade-offs, (3) using backpropagation also for inference to refine existing outputs, (4) deploying learning-to-rank methods for class imbalance. During the period, I contributed to workshops, the Summer School on Computer Vision (VISUM), and other events, and was twice awarded "outstanding recognition" for organizing workshops and helping with the HPC infrastructure.
2014/09–2014/12
Research Grant on Mathematical Modelling Research Mathematics Center of the University of Porto (CMUP)
Epidemiological models for HIV: differential equations, stochastic simulations, cellular automata.

Publications

h-index:
Google ScholarScopusWeb of Science
754

Sources for the following metrics: • Impact Factor (IF) as reported by the journal's webpage. • SJR rank quartiles are from Scimago and relate to the subject category closest to machine learning (not necessarily the best quartile). • CORE rank is from ICORE for whatever last year is available for that conference. • Citation counts come from Crossref. Last update: 2024-11-09

Supervisions

MSc Dissertation

current
João Ricardo Ramos Alves, "Self-supervised Occupancy Networks in Autonomous Driving (Bosch internship)", FEUP
current
Luís Paulo da Rocha Miranda, "Unsupervised Active Learning: Which Frames are Most Important in Autonomous Driving?", FEUP
current
Bruno Vieira Dias, "Trust or not to trust: When to trust the label prediction", FEUP
current
Alexandre Ferreira Nunes, "Semi-supervised Learning on 2D Projections for Autonomous Driving", FEUP
current
Sofía Lucía, "Ordinal losses for range estimation in autonomous driving", FEUP
2024
Airton Tiago, "Data Augmentation for Ordinal Data", FEUP (with co-supervisor: Jaime Cardoso)
2023
Alankrita Asthana, "Iterative Inference for Point-Clouds", TUM
2023
Rafael Cristino, "Introducing Domain Knowledge to Scene Parsing in Autonomous Driving", FEUP (with co-supervisor: Jaime Cardoso)
2022
Pedro Silva, "Human Feedback during Neural Networks Training", FEUP (with co-supervisor: Tiago Gonçalves)
2022
Ana Bezerra, "Phishing Detection with a Machine Learning Net (Internship at E-goi)", FCUP (with co-supervisor: Joaquim Costa)

BSc Project

2024
João Monteiro, "Cross-vehicle collaboration using RGB cameras", FCUP (with co-supervisor: Celso Pereira)
2024
Diogo Mendes, "Automatic Recognition of Pig Activity in an Intensive Production System", FCUP (with co-supervisor: Nuno Lavado)
2024
Beatriz Sá, "Research on Deep Augmentation for Ordinal Regression", FCUP (with co-supervisor: Jaime S. Cardoso)
2024
Eliandro Melo, "Resource Efficiency using Deep Q-Learning in Autonomous Driving", FCUP
2024
Ivo Duarte Simões, "Resource Efficiency using PPO in Autonomous Driving", FCUP
2023
Diana Teixeira Silva, "Condition Invariance for Autonomous Driving by Adversarial Learning", FEUP
2022
Diana Teixeira Silva, "Semantic Segmentation in Neural Networks using Iterative Visual Attention", FEUP (with co-supervisor: Tiago Gonçalves)
2022
Filipe Campos, Francisco Cerqueira, Vasco Alves, "Mobile App using Object Detection for Car Driving", FEUP
2022
Bruno Gomes, Rafael Camelo, "Internship at ANO", FEUP

Internship

2024 July
Isabel Gomes de Sá, "Making Sense of Ordinal Images Without Labels", INESC TEC