The future of radio

AI promises to revolutionize radio by offering personalized content and streamlining operations. However, this technological embrace raises concerns regarding the authenticity and bias of AI-generated content.

An Individual for Customizability

Each of these fictional people is based on the identical patterns and characteristics of a real starting person, ensuring remarkable consistency in the simulated interaction patterns.

2D to 3D

Generation of three-dimensional replicas of real people through a synergistic application of photogrammetry and LiDAR scanning techniques. A high degree of realism was achieved through the accurate reproduction of material textures and the texture of surfaces, complemented by a lifelike light simulation.

A "DNA" as a Constant for Voice and Expression

The uniqueness of this approach is manifested in the fact that the entire performance (dynamics of movement and expression) as well as the vocal characteristics of all these mutations were derived or mutated from a single real person. This base served for further enrichment and personalization of the virtual characters.

The sound was also artificially mutated through AI according to the people to give the different faces the appropriate tone.

Motion Capture via AI

Advanced motion capture methods enabled the precise extraction and transfer of the facial expressions of this person to the 3D models, using widely available smartphone technology.

Artificial intelligence algorithms were specifically used to expand the mimetic subtleties, with particular emphasis on minimizing the so-called "Uncanny Valley" effect – a reaction of eeriness towards almost, but not completely human-like figures.

This question arises, as all too often in the media landscape: Who is really speaking to me?

*Translated by Chat GPT