About me

I’m Andrea Pilzer, a Solutions Architect at NVIDIA based in Italy, focusing on generative AI training and inference. Previously, I worked on uncertainty estimation at Aalto University and domain adaptation at Huawei Technologies Ireland. I completed my PhD at the University of Trento, researching stereo matching, depth estimation, and deep learning.

Research

My research spans computer vision, deep learning, domain adaptation, depth estimation, and uncertainty estimation.

I am a co-organizer of the UNCV (Uncertainty Quantification for Computer Vision) workshop series, that we propose annually at major computer vision conferences.

I regularly review/reviewed for major machine-learning and computer-vision venues, including NeurIPS, ICLR, CVPR, ECCV, ACM Multimedia, and IEEE Transactions on Multimedia.

View publications on Google Scholar ↗

Beyond work

I grew up in the very steep and beautiful Cembra Valley in the Dolomites, Italy.

I have always loved sport, from cross-country and downhill skiing to curling. With friends, I formed Team Trentino curling. We became Italian men’s champions several times, represented the Italian national team at European and world championships, and competed at the 2018 Winter Olympics in PyeongChang.