When we talk about people moving in digital space, this no longer just means communicating on a computer or smartphone via a keyboard. The way that people perceive and interact in the digital world is in a state of upheaval. New technologies are being developed that allow people to be digitalised in 3D so that their digital doppelgangers (digital humans) can be viewed in virtual reality applications. These digital humans are above all interesting for immersive telepresence: virtual worlds that give the users the convincing feeling of being a real part of them.

In augmented or virtual reality, digital humans can see each other and interact. Possible applications include, for example, immersive telecommunication, virtual mirrors in e-commerce and entertainment such as movies or computer games. The particular challenge here: a convincing, immersive experience requires a high-quality reproduction of the depicted person’s appearance and subtle movements.

In the ERC-funded project ‘Learning Digital Humans in Motion’, Thies and his team are concentrating on AI-based image processing and graphic aspects of capturing humans. The researchers will analyse how people move and develop data-driven motion synthesis methods. The aim is to create a sociable digital human with commercially available hardware. In doing so, the question of whether natural language can be used to reconstruct, depict and model the appearance, movement and interactions of digital humans should be answered.