About Me

About Me

Throughout my life, I have been passionate about a wide variety of topics and I can only assume (and hope) this will never change. Now, after around 15+ years of messing around, I understand that I am mostly drawn by three areas: technology, human behaviour, and music. I sometimes manage to combine the first two, whereas the last one has been fairly separated for the rest (until recently, with my exploration of electroacoustic music). In this page, I try to organize some of the things I have done or I am still working through.

For a more professional look on my working career, check my LinkedIn page.

Psychology + Artificial Intelligence

My interest in psychology started in my late teens, after discovering the experiments of Michael Gazzaniga about the split-brain syndrome and getting blown away by all kind of crazy neurological syndromes humans would get when messing around with their brain. This led me to pursue a BSc in Clinical Psychology at the Second University of Naples (which has now changed its name to the mouthful Universita' degli studi della Campania Luigi Vanvitelli). For my thesis (2012) I collaborated with the Laboratory of Autonomous Robotics and Artificial Life (part of the National Research Council) in Rome, under the supervision of Domenico Parisi, producing a work about the Psycho-Sociological Role of Human Storage Through A Multi-Agent Simulation Approach. Basically we simulated a population of agents endowed with an artificial neural network and with different types of storages (public, family, individual) in an environment with scarce resources and see what happened. In fact, interesting things happened, and the agents developed some unexpected behaviours1In the simplest simulations, agent without storage were forced to consume a food token straight away, which sometime would result in a waste of energy if their energy level was already high enough (whereas agents with a food storage would collect the food token and consume it when appropriate). The evolved solution for the agents without storage was this one: once a food token was found in the environment, they would just wait for their energy level to come down before eating it. They would just wait in front of the food token, or they would circle around it in a shark-like fashion. Smart, isn't it? More complex simulations involved different storages and the evolution of altruistic behaviour (sharing the food token with other agents). The agents evolved the behaviour of sharing their food only with genetically related member of the community (parents, siblings, offsprings). They did not develop any altruistic behaviour when the storage was shared across the whole community. .

The platform we used for performing simulation was a heavily modified version of Evorobot. Two different populations are shown here. Different colours represent different gene pools. In the left, the population is very heterogeneous; on the right, they mostly have the same genes, meaning they all come from the same ancestor.

Psychology + Decision Making

After my BSc, I moved to the UK to pursue a PhD in Psychology (2012-2016) at the University of Plymouth under the supervision of Chris Harris. After chasing different possible threads I got interested in exploring optimization strategies in humans in those cases where they had to trade speed with accuracy. I believe my main contribution to the field is the EXACT paradigm, in which the speed-accuracy trade-off is made explicit on the screen in front of the user, hugely simplifying the understanding of what decision strategy was used instead of having to make a great number of assumption about how the brain works internally. In other words, the EXACT paradigm is an abstract-version of a classic psychological experiment (which is already an abstract version of real-life scenarios). If this sounds very vague, it's because I would need many more words to explain it properly, so just go and read the paper or contact me for full details. I still strongly believe in this paradigm, but unfortunately, this had basically zero resonance with the researcher field. Oh well, maybe it will emerge in the future and be a big hit!

Computer Vision + Virtual Reality + Machine Learning

During my PhD in Psychology, I got interested in Machine Learning and found a job at Ultraleap in 2016 (which at the time was called Ultrahaptics) in Bristol. As my first experience in industry, this was great fun. During the first year my work consisted of building from scratch (and then maintaining and updating) a gesture recognition system which would allow the user to record a gesture which would be then detected and classified. This hand-gesture was detected by an infrared camera, the Leap Motion (which has now joined forces with Ultraleap - at the time it was a totally independent company). I also worked on different demos, expecially in Virtual and Augmented Reality, using Unity as the main platform. Example, this project:

No, I am not one of the talking heads in the video, but at least I got to be in some of the B-roll! Yes, it's me at 1:50!

I then worked as part of the tracking before Leap Motion joined us. Most of this stuff was under heavy NDA so I would just say that during that time I greatly expanded my knowledge of Deep Learning, Python, Tensorflow, and all the happy Machine Learning family.

Music

I have played the piano since I can remember. I got my Diploma in Piano Performance (10/10 with honours) in 2008 at the Salerno Conservatoire G. Martucci, with a particular focus on Baroque music. After that, I had to stop playing for a decade due to back pain emerging just when playing the piano2Which I now believe to be a result of sensitization and possibly trigger point. I did, in the meantime, become very interested in back pain and in general chronic muscular condition such as Fibromyalgia and Chronic Fatigue Syndrome. Overall, I can now play and compose without much problem, but my journey fighting chronic back pain has not finished - and probably will never finish. In 2017 I started playing again and realized that I couldn't be bothered to be a bad cover band of The Great Composers From The Past™, and I started being a bad cover band of my own pieces while working as an Engineer in Bristol (see above). After messing around with electronic music, I started to consistently and uniquely compose for piano. I performed my pieces in several venues, both in UK and in Italy. Two highlights:

I keep writing and performing around and keenly exploring new directions, currently focusing on integrating electronic elements in my work (but none of that stuff has seen the daylight yet).

Miscellaneous

I really liked working with Unity for my job in Bristol, so I started messing around with it in my free time as well. This resulted in Physics Scatter, a Unity Tool for placing object around in your virtual world in a realistic fashion, following physics.

Automatic Page Turner For Pianists is supposed to be a solution for pianists who wanted to turn their pages without having to lift a finger - mainly because their fingers were already busy playing terribly difficult stuff. It was a random old project that I would eventually develop into an app and become filthy rich.