Antonio Carta

Author Image

I am an assistant professor at the Department of Computer Science of the University of Pisa, working with the Computational intelligence and Machine Learning group (CIML) and the Pervasive AI Lab (PAILab). I am also a member of ContinualAI and the lead maintainer of Avalanche, a continual learning library.

My research is focused on continual learning methods applied to deep learning models. I work on computer vision and sequential domains (natural language, time series). In my opinion, continual learning is one of the key elements to design scalable and efficient deep learning models.

Recent Research

Ex-Model: Learning from a Stream of Trained Models A novel scenario where CL models learn from domain experts without sharing any real training data (preprint, post).

Continual Pre-Training Mitigates Forgetting in Language and Vision We study the continual training of large models and show that, under appropriate conditions, they are robust to catastrophic forgetting (preprint).

News

(2022-03-31) We are organizing the 1st International Workshop on Pervasive Artificial Intelligence hosted by the 2022 IEEE World Congress on Computational Intelligence. Submission deadline: 2022-04-30 (webpage)

(2022-02-01) I just started a new position as an assistant professor (RTD-A) at the University of Pisa.

(2022-01-01) The CLVISION workshop on continual learning has been accepted @CVPR22.