Oriol Pujol Vila

  • Vice-President for Digital Transformation
  • Universitat de Barcelona
  • oriol_pujol@ub.edu


My research is focused on the design of Machine Learning algorithms. In particular supervised learning optimisation strategies for large scale data sets.


I teach "Agile Methodologies for Software Development", and "Machine Learning" at the Fundamental Principles of Data Science master's programm and "Introduction to Machine Learning" at the Artificial Intelligence master's joint programm involving UPC, URV, and UB.


I am currently the vice-president for Digital Transformation at Universitat de Barcelona and Principal Investigator of the Vision and Computational Learning consolidated research group (SGR).

Short Bio

Oriol Pujol Vila is tenured associate professor in Computer Science and Artificial Intelligence at the department of Matemàtiques i Informàtica at Universitat de Barcelona. He obtained the degree in Telecomunications Engineering in 1998 from the Universitat Politècnica de Catalunya (UPC). The same year, he joined the Computer Vision Center and the Computer Science Department at Universitat Autònoma de Barcelona (UAB). In 2004 he received the Ph.D. in Computer Science at the UAB with a work in deformable models, fusion of supervised and unsupervised learning and intravascular ultrasound medical image analysis. In 2005 he joined the Dept. of Matemàtica Aplicada i Anàlisi at Universitat de Barcelona (UB) where he became tenured associate professor. He currently leads the Vision and Computational Learning consolidated research group (SGR). He has published more than one hundred and fifty articles in machine learning, computer vision, and their applications. He has more than eighteen years in knowledge transference in data analysis projects is fields such as finance, health, marketing, wearable sensors, among others. He served as director of Computer Science undergraduate studies, director of the postgraduate courses on Data Science and Big Data, and director of the official master's program in Fundamental Principles of Data Science. He is currently vice-president for Digital Transformation at the University of Barcelona.


Research highlights

Supervised Online Learning Algorithms

My current basic research lines are focused on the topic of online learning optimization strategies. My current goals are to design new optimization algorithms that allow machine learning techniques to work with low computational resources and to propose new computational models in the task of supervised learning.

Deep Learning

My current research line in deep learning is focused in designing architectures that allow the use of weak supervisory signals for efficient learning of deep learning models in the presence small data sets. I am also interested in generative algorithms and the role of uncertainty for sampling plausible instances.

Ensemble Learning

One of the topics I have been researching for a long time is ensemble learning. In particular Error Correcting Output Coding techniques.

Applications of Machine Learning

My main area of application is computer vision, though I have applied machine learning methods in many other domains, such as finance, click-through-rate prediction, e-health, or medical imaging among others.


Machine Learning

I am currently teaching Machine Learning at Foundamental Principles of Data Science master's program and also at the Artificial Intelligence master's program. I like to use Jupyter Notebooks and live coding sessions for teaching this subject. My goal is that students not only have the basic understanding of the subject but also know how to effectively apply it, and furthermore, are able to code from scratch most of the most well known machine learning techniques, e.g. support vector machines, deep learning algorithms, random forests, among others.

Software Engineering and Agile Methodologies

I teach agile methodologies for undergraduate Computer Science students. This includes Scrum, Kanban, and Lean principles. Besides these subjects basic skills for leadership, project, team, and time management are also taught. Two particular subjects I like to emphasize in this course are mediation and negotiation.

Previous teaching: Image Processing

I taught Image Processing and Computational Photography.

Selected Publications

Check my Google Scholar profile for a comprehensive list of my publications

  • Deep Learning

  • Basic Algorithms for Machine Learning

    • Approximate polytope ensemble for one-class classification

      P Casale, O Pujol, P Radeva, Pattern Recognition, 2014

    • Multi-scale stacked sequential learning,

      C Gatta, E Puertas, O Pujol Pattern Recognition, 2011

    • Geometry-based ensembles: toward a structural characterization of the classification boundary,

      O Pujol, D Masip, IEEE Transactions on Tattern Analysis and Machine Intelligence, 2009

  • Error Correcting Output Coding

    • Error-Correcting Factorization,

      MA Bautista, O Pujol, F De la Torre, S Escalera, IEEE Transactions on Pattern Analysis and Machine Intelligence, 2017

    • On the decoding process in ternary error-correcting output codes

      S Escalera, O Pujol, P Radeva IEEE Transactions on Pattern Analysis and Machine Intelligence, 2010

    • Subclass problem-dependent design for error-correcting output codes

      S Escalera, DMJ Tax, O Pujol, P Radeva, RPW Duin IEEE Transactions on Pattern Analysis and Machine Intelligence, 2008

    • Discriminant ecoc: A heuristic method for application dependent design of error correcting output codes

      O Pujol, P Radeva, J Vitria IEEE Transactions on Pattern Analysis and Machine Intelligence, 2006.

  • Applied Research

    • A Gesture Recognition System for Detecting Behavioral Patterns of ADHD

      MA Bautista, A Hernández-Vela, S Escalera, L Igual, O Pujol, J Moya, V Violant, MT Anguera, IEEE transactions on cybernetics,2016.

    • Personalization and user verification in wearable systems using biometric walking patterns

      P Casale, O Pujol, P Radeva Personal and Ubiquitous Computing, 2012

    • Human activity recognition from accelerometer data using a wearable device

      P Casale, O Pujol, P Radeva Iberian Conference on Pattern Recognition and Image Analysis,2011

    • Rayleigh mixture model for plaque characterization in intravascular ultrasound

      JC Seabra, F Ciompi, O Pujol, J Mauri, P Radeva, J Sanches IEEE Transactions on Biomedical Engineering, 2011

    • Traffic sign recognition using evolutionary adaboost detection and forest-ECOC classification X Baró, S Escalera, J Vitrià, O Pujol, P Radeva IEEE Transactions on Intelligent Transportation Systems, 2009

Knowledge transference

Research projects, applied research, and knowledge transference

Health and medical applications

Medical Imaging, Computer Assisted Diagnosis

Wearable systems

User Authentication, Action Recognition

Civil and Industrial applications

Traffic Sign Recognition, Industrial Vision

Online marketing and finance

Click-through Rate, Time Series Forecasting