Extending the Summer with the Help of AI: Exploring the Mediterranean Machine Learning Summer School

10/14/2024

The Mediterranean Machine Learning Summer School (M2L), held in sunny Milan in September 2024, was a week packed with cutting-edge AI research, insightful talks, and plenty of networking opportunities. Organized and sponsored by Google, Google DeepMind, and Reply, this event attracted top-tier researchers and budding talent from around the globe. Among them was Francesco Di Salvo, a PhD candidate from xAILab Bamberg, who showcased his work on "Privacy-Preserving Datasets by Capturing Feature Distributions with Conditional VAEs."

A Week of AI Insights

The M2L Summer School was filled with talks from top scientists representing Google, DeepMind, Mistral AI, the University of Oxford, and more, showcasing the latest breakthroughs in Language Models, Computer Vision, and Reinforcement Learning.

On top of that, the daily poster sessions provided a great platform for attendees to present their own work. It was a collaborative environment where master’s students, PhD candidates, senior researchers, and industry experts exchanged ideas and sparked new collaborations.

Francesco's poster on privacy-preserving datasets was a highlight, showcasing how Conditional VAEs can be used to capture feature distributions while safeguarding data privacy. His research was well-received, and he’ll be presenting it again at the British Machine Vision Conference (BMVC) in Glasgow this coming November. For more details, you can access the paper here and the associated code repository here.

The summer school was not just about learning—it was about connecting with a global community of AI enthusiasts and professionals, making it an unforgettable experience for everyone involved.