About me

Bio

I was born in Soria, a small city in central Spain, and grew up near Madrid. In 2009 I started my bachelor degree on Audiovisual Systems Engineering at the University Carlos III of Madrid, during which I had the chance to study one year as an Erasmus fellow at the University of Stuttgart, Germany. In 2013 I obtained a research fellowship to complete my bachelor thesis at the Group of Multimedia Processing (GPM) with Fernando Fernández Martínez and Fernando Díaz de María. At the GPM I worked as a research and teaching assistant during my master studies, working on projects related to image processing and computer vision. The main topic of my research was the automatic recognition of aesthetics and elicited affect of natural videos.

In February 2016 I was selected as a Marie Skłodowska-Curie PhD fellow at Prof. Peter König’s lab at the Department of Cognitive Science of the University of Osnabrück and at EyeQuant, in Berlin. Being part of the Innovative Training Network NexGenVis gave me the opportunity to collaborate with a great network of PhD students at leading laboratories in the field of visual neuroscience. During my PhD, I was a visiting PhD student at at the Spinoza Center for Neuroimaging in Amsterdam with Dr. Serge Dumoulin and at the Cognition and Brain Sciences Unit of the University of Cambridge with Dr. Tim Kietzmann. In November 2020, I defended my PhD thesis, and in December I moved to Montreal to start as a postdoc at Mila - Quebec Artificial Intelligence Institute, with Prof. Yoshua Bengio.

Research

My main research focus is on brain-inspired deep learning and computational neuroscience. In particular, I have studied data augmentation as a biologically and perceptually informed way of training convolutional neural networks, compared to explicit regularization techniques such as weight decay and dropout. Besides, I have collaborated in studies of visual perception and visual attention involving eye-tracking and analysis of fMRI data. At Mila, I am currently working on projects to fight climate change using machine learning, such as an educational tool to visualise the effects of climate change on real photos.

Besides my actual research, I am very interested (and concerned) about the way we scientists do science, what some people call meta-research. I am a strong proponent of open science, in its broadest sense: I care about reproducible, rigorous and transparent research, but also believe that science (and everything, in general) should be inclusive and accessible, without leaving any minority or less privileged groups aside. Moreover, I advocate for a more cooperative, less competitive, way of doing research (and everything, in general).

Some specific actions I have taken towards these goals are: make most of my data and code available in open-source repositories, as well as sharing educational resources that may be useful for other people, such as this repo, with some Linux configuration files and tips about tools; this repo with examples of how to perform bootstrap with Python; and this repo with the source code of my PhD thesis in LaTeX. I also try to be active on StackOverflow and similar platforms. In an attempt to contribute to more accessible science, I am committed to promote publication venues (conferences and journals) that are open-access and have low registration or publication fees. Specifically, I have decided not to submit first-author manuscripts to for-profit organisations, and I try to encourage the same policy when I am not the first author. In the same spirit, I am building a list of publication venues in both machine learning and neuroscience, where I highlight their open-access policy and fees, as well as whether they are run by not-for-profit or for-profit organisations.

Finally, I joined this initiative to build a list of machine learning mentors to help mitigate systematic discriminations in academia. If you consider yourself a member of a discriminated group and think that I could help you with something (for example, reviewing a manuscript, application, CV, etc.), please do not hesitate to get in touch, without any need to justify in which way you consider yourself discriminated against.

Beyond research

When I am not analyzing data, coding or reading papers, I like spending time with my friends, doing things as varied as having a conversation over a beer or a cup of coffee or tea, playing music, reading books, climbing, going to punk-rock concerts and festivals, travelling…

If I had a second life, I would like to be a musician at a symphonic orchestra. I guess it is because I have played the violin since I have memories. Fortunately, this life has also given me the chance to play the violin at several orchestras and in a band in Berlin, The Broken Jug Ramblers, among many other unforgettable musical experiences. I also played the bagpipe with a big pipe band from Madrid, Lume de Biqueira, which was a lot of fun.