I am a Ph.D. candidate in the Signal Processing Applications Group, attached to the Information Processing and Technology Center, at the Universidad Politécnica de Madrid under the supervision of Dr José Luis Blanco. My doctoral studies specialize in the field of audio generation, with a specific focus on leveraging generative AI techniques for producing high-quality audio effects. My research aims to capture the full variability of reality in procedural audio effects, contributing significantly to the realms of film, video games, and various other industries.
My investigations have extensively explored the use of variational autoencoders in this context. I have developed autoencoders capable of processing audio phase, implementing them for web browser usability. This research has resulted in multiple publications in the Journal of the Audio Engineering Society (JAES).
Moreover, my knowledge of applied artificial intelligence in this domain has enabled me to contribute to Spain’s first short film employing audio effects on Foley sounds. This “El Testigo” project emerged as a practical experiment of my expertise. I provided the director with a tool I designed named Foley-VAE, which allowed for exploring and creating various novel sounds. “El Testigo” will soon be released on multiple platforms.
My doctoral studies afforded me the unique opportunity to visit one of the world’s premier research centers in this field: the Centre For Digital Music (C4DM) at Queen Mary University of London. During my stay, I collaborated with Joshua Reiss on a new area: articulatory speech production. We optimized a human speech model for generating non-speech vocal sounds (yawns, laughs, cries, etc.). This research is particularly intriguing as it addresses a highly demanded effect in the film industry and represents a significant paradigm in audio generation. Integrating vocal tract information into AI is an exciting frontier, leading us to present at DAFx, a prestigious conference in Copenhagen. This period was not only professionally rewarding but also personally enriching, thanks to the wonderful people I met, including David, Vjosa, Elona, Nelly, Mónica, Zhiyuan, Yisu, Pedro, Jack, Saurjya, Guille, José, Adam, Alex, Bleiz, Ben, Chris, Christian, Fran, Jordie, Louise, Brendan, Drew, and many others.
My interests focus on real-time sound transformation for real and virtual conditions. Specifically, the adaptation and production of sound effects to certain types of events and environments to give a more natural feel to the sound. I am interested in production adaptation for FX and events, emulation/auralization of virtual environments, context tagging, and audio mixing.
I currently hold a teaching assistant position at the university. I give classes in subjects related to statistics and deep learning. I am also a tutor for several master’s theses.
On the other hand, I am passionate about entrepreneurship. In 2020 I launched with other colleagues an artificial intelligence startup applied to business geointelligence: Pickgeo. I like to consider this experience as a triple master’s in law, management, and software development in one, resulting in a thriving company.
On this website, you will find information about me and my work. Maybe also about some relevant personal experience.
Keep reading? Have a nice day.