Dr. Xavier Porte received the MSc degree in 2011 and the Ph.D. degree in 2015, both in Physics, from the University of the Balearic Islands (UIB), Spain. His MSc and Ph.D. main research topic was nonlinear dynamics of delay-coupled semiconductor lasers. From 2016 to 2018 he was with the Technical University Berlin, Germany, where he was engaged in research work on an external optical coupling of quantum-dot microlasers for nanophotonic applications.
He is currently a postdoctoral researcher at the FEMTO-ST Institute in Besançon, France, working on photonic spatio-temporal neural networks. His research interests range from nonlinear delay dynamics to nanophotonics and neuromorphic computing. The title of his MULTIPLY project is “Networks of autonomous all-photonic neurons”. Xavier told us more about special place FEMTO-ST and his research.
Is your host institution, FEMTO-ST, a special place for you?
There are several things that make FEMTO-ST a special institute. I would say the most important ones are its location and the unusual combination of many advanced scientific and engineering disciplines. First, FEMTO-ST is not placed in a big Capital city like Paris or London. It is instead located in a beautiful medium-size city named Besançon, which I recommend everybody to visit as it has a UNESCO human heritage site right in its center! Besides that, the whole region of Franche-Comté is full of amazing nature and lots of options for hiking and other outdoor sports. As a research site, what makes FEMTO-ST so special is that it gathers cutting-edge research from several different disciplines, not only optics. You can find groups doing advanced micromechanics, robotics, nanofabrication, metrology, etc. It is a truly rich collaboration environment!
Has your area of interest changed or expanded due to the work on the MULTIPLY project?
Definitively! My area of interest has tremendously expanded during my postdoctoral fellowship. MULTIPLY has offered me a unique platform from which to discover and interact with other research fields in photonics. Particularly fruitful are the summer schools and thematic workshops organized by MULTIPLY, with speakers from several different fields but focusing on a specific topic. These are great opportunities to meet the interdisciplinary community in MULTIPLY.
Did you find any interesting unexpected collaborations during your MULTIPLY project?
My MULTIPLY project is devoted to the implementation of a hardware-based autonomous photonic neural network. In our group, we face this challenging task from scratch, so it requires every bit of invention and not rarely to seriously consider a-priori foolish ideas. So yes, during the last two years and under the guidance of Dr. Daniel Brunner, I had the opportunity to meet several collaborators from very unexpected fields and this really boosted my scientific perspective.
FEMTO-ST research group
What is innovative about your research?
I think what is really innovative about my research is the aim to invert the perspective on how we look at machine learning. Artificial neural networks represent an extremely successful paradigm and the most natural action would be to simply profit from them to improve the design and performance of some system. But I don’t do that, I rather revert the point of view and design systems that naturally perform well when operated as an artificial neural network. This change of paradigm may potentially result in more efficient, fast and performing neural nets.
What do you want to achieve with your research?
A complete photonic toolbox for neuromorphic computing. I want to recreate the different constituents of artificial neural networks with photonics systems and implement their functionalities leveraging their optical properties.
Why is being a scientist cool?
In one sentence: because of the pleasure of finding things out. Everyone can choose how to See the world, either with blind acceptance and based on dogmas or with a critical mindset and based on logic. Science naturally follows the second option.
