Computer Vision News - June 2023

27 Learn2Learn (L2L) of data we have and then be adapted with very few examples on a new domain, ” Stefano tells us. “ That’s what we want to know through the challenge. Through many people trying out their methods and ideas, we hope to get a better benchmark than just us in our group thinking about this and benchmarking a few methods. ” To learn more about cross-domain few- shot learning and the background to this challenge, watch excerpts from our video interview with Stefano and Susu. Of course, this is a medical challenge, so the medical imaging community is invited to join. However, Stefano and Susu also expect participation from scholars in non-medical machine learning fields , as few-shot learning and meta-learning are currently popular topics in machine learning research. The challenge has been designed to be easily accessible to machine learning researchers from any domain. “ One thing we were trying to do ourselves is to find out the best method to build an algorithm that can learn from the plethora