Bio

I’m a PhD Student researching Deep Learning at Politecnico di Torino in the Vandal lab. I’m passionate about new technologies as I’m convinced they can improve our life. Still, I firmly believe these improvements should benefit all: knowledge should be open, personal data should stay private.

Download my complete résumé.

Skills

I speak Italian, English and French.

I am a skilled developer and a fast learner. I have used Linux for my whole life and thus have some sysadmin experience.

I’ve also got some experience in many of the most common programming languages as C, C++, Python, Java, and Bash, but I have also used SQL, Javascript, and PHP.

I have developed native applications for Linux with C/C++, GUIs with Qt, and mobile applications with Android Studio. I used scikit-learn for traditional Machine Learning problems and Hadoop and Spark for Big Data analysis.

I develop Deep Learning projects with PyTorch, Jax+Flax and any other tool that can come in handy.

I use Arch Linux and vim, because I love simplicity.

Research interests

It is well known that Deep Learning solutions have enabled to obtain excellent performance in many Computer Vision tasks in the last years. Still, the transition from lab settings to real world applications is not trivial and can, in some cases, determine a drastic reduction in the effectiveness of some innovations. My research goal in the long term is to alleviate some of the causes of this performance drop, in particular I focus on semantic and visual domain shifts. With this objective in mind, I work on the design of models which prove a great robustness to test scenarios that differ from the training ones. This result can be even more interesting if it is reached in label-scarce settings, i.e. without requiring human intervention to guide models learning. The desired result is to build models that can be applied in visual domains for which no data or no labels are available at training time, without recording a significant performance drop, and at the same time that are able to correctly discern between objects they know and objects not seen before.

I generally focus on the tasks of visual object detection and object recognition.

Personal interests

I come from a small town in the mountains and I really love hiking!

I also like cooking and reading, I used to think that one day I would become a writer!

I believe that Climate Change is the most important battle of our time.