soundwalking deep latent spaces

This research-creation project explores soundwalking as critical and embodied practice of AI and deep latent spaces.

Soundwalking consists in an individual or collective walk with a focus on listening to the environment. It has joint pedagogical, political and artistic implications, including educating persons to listen to uncanny sounds, giving voice to places and bodies that have been muted throughout history, or articulating musical performances throughout such situated soundscapes. We believe that soundwalking could help reveal social, political and planetary issues raised by modern AI applications, while suggesting new practices of deep generative models.

We created a soundwalk experience where a human person moves through a three-dimensional virtual world to explore a latent sound space generated by deep learning. We crafted a deep learning prototype leveraging nn∼ and New Atlantis as computational platforms for AI audio processing and virtual world experimentation. These early experiences let us look at music AI against latent soundwalk and deep latent spaces against virtual worlds. We will pursue our art-research collaboration with the hope to contribute to community-wide reflections toward new AI for musical expression.

This slideshow requires JavaScript.

The project is being developed with Ludmila Postel, in collaboration with the Locus Sonus Vitae group of ESAAIX, in the context of a postdoctoral fellowship program supervised by Emanuele Quinz at EUR ArTeC (Université Paris 8), in collaboration with the Reflective Interaction group of EnsadLab.
Talk @ Resounding the Deepscape, Paris, Fr (Dec.14.2022)
Paper at NIME 2023