Skip to main content
European Commission logo
European Innovation Council and SMEs Executive Agency (EISMEA)
News article23 September 2022European Innovation Council and SMEs Executive Agency

SeerPredict: A smarter AI camera for live video productions

The Horizon 2020 INNOSUP-funded project SeerPredict delivers a smarter AI camera that elevates automation for lecture capturing, corporate events and stage productions.

SeerPredict
©Seervision, 2022

The EU-funded project SeerPredict, by Swiss deep-tech startup Seervision, delivers smarter AI cameras that autonomously film talent during live video productions - just like a camera person would. They combine novel machine learning algorithms with an innovative robotics and control platform to robustify Seervision’s automation solution and improve its performance in scenarios where people are crossing or occluding each other.

SeerPredict innovates camera automation solutions for video productions

The demand for video content has been increasing rapidly for the past decade. This increased demand cannot be met with human operated cameras, which are expensive, harder to find and bring extra administrative complexity to a video production. As a result, the whole industry is seeking solutions to automate all aspects of video productions, while maintaining high quality and reliability standards. Seervision’s solution, which is able to reliably operate during crossings and occlusions, will help address more advanced use cases.

SeerPredict combines state-of-the-art machine learning algorithms with an advanced robotics and control platform. The resulting improvements deliver autonomous visual tracking that is 2x faster, more reliable and that moves cameras to keep video production talent framed perfectly - just like a camera person would. Moreover, the developed solution reduces the required computational resources by about 50% and paves the way for further improvements. This helps Seervision to better serve new and existing customers across use cases in education, corporate and live events.

SeerPredict helps reduce computational resource requirements

As a result of this Horizon 2020 INNOSUP-funded project, Seervision has been able to further improve the responsiveness and reliability of its solution. More importantly, due to the reduction in computational resource requirements, with the 41.0.0 “Susten” release in August, 2022, it is now possible to support two AI cameras in one compute unit instead of just one, realising a 2x hardware cost saving and rack space reduction for Seervision’s customers.

For more information on their project, visit Seervision’s website.

Details

Publication date
23 September 2022
Author
European Innovation Council and SMEs Executive Agency