As machine vision and robotic automation evolve towards new levels of intelligence, broadcast production systems will advance as well. This will pave the way for new approaches to production, with multi-camera robot and drone setups collaborating to segment and understand unexplored scenes, and then executing all of the operational tasks handled by traditional camera crews in the past – leaving creative storytelling to humans.
At Wyss Zurich, the Seervision team of engineers, programmers and broadcasters is developing a radically new technology called “adaptive motion control.” The project is the result of years of research at ETH Zurich’s automatic control laboratory, one of the world’s top three laboratories of this kind.
The software, which is both camera- and robot-independent, incorporates image recognition, artificial intelligence, model predictive control techniques and high-speed dynamics to make autonomous video production possible. Seervision has analyzed the motion sequences of cameramen working in a variety of applications and converted them into algorithms that allow for real-time object recognition and scene segmentation. Moreover, these algorithms are continually enriched through a machine learning process, resulting in multiple robot camera setups in which each robot can not only autonomously perform all the tasks of traditional camera work, but also exchange information with other robot cameras. The final product is indistinguishable from what is produced by a team of human beings. However, the production is now controlled by a single operator, who is free to focus creatively on the big picture.
The technology developed by Seervision represents a revolutionary change in studio and sports production. It makes it possible to achieve consistently high quality with significantly lower production costs and allows for innovative production techniques.
The team’s mission is clear: to ensure that in the future, every robotic camera system will use Seervision’s adaptive motion control software.
Download Seervision fact sheet