online interactive installation, facial recognition, AI-generated sound

Collaboration with Nikita Prudnikov, AI-expert, musician
Faces2Voices is an online interactive installation which uses facial recognition technology and AI-synthesized sound to create a generative music composition based on imaginary voices of online visitors. The composition is evolving in time depending on the contributions of people involved.

Lots of governments use surveillance technology as a way to control the spread of COVID-19. At the same moment many citizens are ready to give up some privacy for the common good. But how can we define what level of privacy should we give up and how much data do governments really need to respond effectively?
We focus on critical approaches to AI technologies and explore these kinds of questions within the procedural rhetoric in Faces2Voices project.

People could contribute to the project by granting access to their device's camera. AI recognised faces, synthesized imaginary voices and added them to the live stream.

website: faces2voices.live

We reflected on Speech2Face