Hello, visitor! 🖖

I am a Data Scientist at Verizon Connect in Florence. I kicked things off at Target Reply in Milan before coming back here to get my Ph.D. in Machine Learning working at the Global Optimization Laboratory.
I'm sticking around Florence and now work at Verizon. My background is a mix of Math and Computer Engineering. These days, I'm really into computer vision, intelligent vehicle applications, and multimodal models. In the past, I've also spent time working on continual learning and sustainable AI.
I know next to nothing about front-end. This page was realized by simply vibing with Gemini 🤖.

Email GitHub LinkedIn Bluesky

Blog

Publications

EFC++: Elastic Feature Consolidation with Prototype Re-balancing for Cold Start Exemplar-free Incremental Learning

S. Magistri, T. Trinci, A. Soutif-Cormerais, J. van de Weijer, A.D. Bagdanov (2025).

[Under Review since July 2024]

[PDF] [Code]

How green is continual learning, really? Analyzing the energy consumption in continual training of vision foundation models

T. Trinci, S. Magistri, R. Verdecchia, A.D. Bagdanov. (2024).

Computer Vision – ECCV 2024 Workshops, Part XXII, pp. 300-317.

[PDF] [Code]

Elastic feature consolidation for cold start exemplar-free incremental learning

S. Magistri, T. Trinci, A. Soutif-Cormerais, J. van de Weijer, A.D. Bagdanov (2024).

The Twelfth International Conference on Learning Representations (ICLR).

[PDF] [Code]

Cross-Model Temporal Cooperation Via Saliency Maps for Efficient Recognition and Classification of Relevant Traffic Lights

T. Trinci, T. Bianconcini, L. Sarti, L. Taccari, F. Sambo. (2023).

International Conference on Intelligent Transportation Systems (ITSC), pp. 2758-2763.

[PDF] [Code]

Patents

Systems and methods for reducing power consumption of executing learning models in vehicle systems

T. Trinci, T. Bianconcini, L. Sarti, L. Taccari, F. Sambo. (2024).

US Patent App. 18/334, 840.

[PDF]