In nuce: A group of eight computer scientists has published Rich – a database containing 142 video recordings in 4K resolution (recorded from various perspectives). The recordings show people interacting with a real environment. The acronym Rich breaks down as follows. R: Real Scenes. I: Interaction. C: Contact. H: Humans. The computer scientists involved in the project are Chun-Hao P. Huang, Hongwei Yi, Markus Höschle, Matvey Safroshkin, Tsvetelina Alexiadis, Senya Polikovsky, Daniel Scharstein and Michael J. Black.
What is special about Rich? According to the developers, Rich is the first time that human-scene contact (HSC) has been successfully generated using a single image. According to the developers, HSC is the first step towards understanding how people interact with their environment. The computer scientists go on to explain that significant progress has now been made in recognising 2D human-object interaction (HOI) and restoring 3D human pose and shape (HPS) – HSC, on the other hand, is an ongoing challenge.
Click further: All further information on the Rich project. In the following video “Capturing and Inferring Dense Full-Body Human-Scene Contact”, the project participants explain how Rich works in general and HSC in particular.
[CVPR 2022] Capturing and Inferring Dense Full-Body Human-Scene Contact
