Learning continually from a never-ending stream of data is a key property for every Pervasive AI system. Being able to sustainably and effectively adapt to the ever-changing environments and circumstances of the real-world is often defined as a key property of Intelligence.

At the PAI Lab we design and implement deep continual learning algorithms for enabling the next generation AI systems and study their applications to real-world problems. In particular, we are interested in:

  • Unsupervised / Self-Supervised/ Weekly/ Semi-Supervised Continual Learning
  • Continual Sequence Learning
  • Neuroscience-Inspired Continual Learning
  • Continual Reinforcement Learning
  • Continual Learning R&D Frameworks & Tools
  • Continual Robot Learning
  • Continual learning on the Edge
  • Distributed Continual Learning
  • Real-World Continual Learning Applications
  • …and much more!

Team

  • Davide Bacciu – Associate Professor
  • Vincenzo Lomonaco – Assistant Professor
  • Claudio Gallicchio – Assistant Professor
  • Antonio Carta – Post-Doc
  • Andrea Cossu – PhD Student
  • Rudy Semola – PhD Student
  • Michele Resta – PhD Student
  • Valerio De Caro – PhD Student
  • Hamed Hemati – PhD Student (co-supervised with Damian Borth at University of St. Gallen)