Tech Insight: Stage Events
A closer look at Seervision’s defining technologies
How to contact us
This series of case studies illustrates how Seervision is disrupting the video production industry and how our system be implemented in support of video productions. If you want to learn more about Seervision and this use case, don’t hesitate to contact us at mail@seervision.com or visit us on LinkedIn, we will be happy to answer your questions.
Stage Events
Welcome to a new post inSeervision’s Tech Insight series. This time, we present our use case for stage events, and we will discuss how Seervision can revolutionize the production workflow, from TED Talks to CEO presentations. We will also share insights behind the technology we have been developing in the past years.
To explain the application of Seervision’s technology in this use case, we start by analyzing the pain points affecting this industry.
In recording a single event, many camera operators are involved, which demands a high level of coordination among cameramen and the editor. A great example would be Google’s I/O 2019 event, where various cameras are framing the speakers in a combination of simple and complicated shots.
Such combinations of repetitive and lengthy shots can be tedious since the cameramen need to focus their attention towards avoiding any possible framing mistakes. The involvement of such skilled camera operators generates high costs, which are not always affordable by the event organizers. This leads to limiting the quality of video productions based on a specific budget. Going back to the example of Google’s presentation, to give a positive impression of the products presented, it is essential to have proper broadcast coverage of the event. This requires a high headcount of professional cameramen and translates into a quality-cost tradeoff. At Seervision, we address all those pain points.
Through our user interface, a single operator can view and control multiple cameras simultaneously by connecting each of these cameras to a Seervision DOP [Director of Photography]. The DOP is, in fact, the “brain” that processes the live image from the camera. According to the operator’s preferences, the DOP gives the relevant instructions to the robotic head moving the camera.
Furthermore, thanks to the possibility to create preset positions, tedious tasks such as a repetitive transition of the camera over the audience can be automated while maintaining a pleasant cinematic look.
Finally, by using our adaptive motion control technology, the speaker on stage can always be kept in the desired framing without any intervention needed by the human operator.
Altogether, Seervision offers a high-quality solution allowing video production companies to ensure an excellent result even with a limited available budget.
Behind the tech curtain
It’s time to hold your breath and take a deep dive in Seervision’s technology to understand the technical background of the features we have discussed above.
One of the critical features of Seervision’s architecture is the ability to remotely control the camera, eliminating the need for the camera operator to be physically next to it.
To understand how the cameraman can be controlling the camera from anywhere in the world, let’s see how the different components of our system communicate with each other.
The camera and the robotic head are connected with the “brain” (the DOP) via a standard SDI cable and an ethernet cable respectively. The connection can be either direct or indirect through the venue’s network and video infrastructure if the venue is set up to support this kind of communication. The user can communicate with the system and the DOP via a specific interface designed by Seervision.
Since the interface is using standardized internet protocol to communicate with the rest of the system and is browser-based, it allows users with sufficient internet connection to control the camera from anywhere in the world.
Furthermore, from a single user interface, it’s possible to control multiple DOP and thereby, multiple cameras. Finally, since our system is designed to integrate seamlessly into existing production workflows, it can interact with existing hardware panels such as Skaarhoj’s PTZ Extreme.
Now that we have seen how the different components of the system communicate, let’s discuss the basic functionalities of the system, which are Manual Control and Presets. These features are particularly exciting since from any status and position of the camera, it’s possible to create or move to predefined positions without interrupting the movement of the camera. The key element is the movement between two positions. Nobody likes sudden, jerky motion when watching a video feed.
That is why at Seervision, we use state-of-the-art control techniques to achieve the best cinematographic performance without compromising the quality of the shot.
Repetition and consistency are some of the key features of the basic PTZ functionality Seervision can offer.
These features allow the system to move the camera just like cameramen by analyzing the movement and using information regarding the model of any Pan Tilt head to predict and adapt to any situation.
Keeping the best for the end, adaptive motion control allows tracking moving subjects in the frame, and it uses state of the art artificial intelligence and neural networks to bring new levels of accuracy to productions while minimizing footprint and production costs. Face and limb detection algorithms are used to analyze the scene and detect human targets. Specifically trained neural network algorithms are used to predict their movement, determine cinematography, and automate camera operation.
In terms of scalability, there is no limit to how many objects can be detected and distinguished, and once the main subject is selected, the system is designed to focus on providing stable, natural-looking content.
At Seervision, we are convinced that AI should support people. Thus, our system allows the operator to take care of the overall picture and leave the tedious work to the AI. At any time, the system also allows the operator to override our algorithms and adjust the picture according to his taste.
Conclusion
To sum up, Seervision’s system delivers in five ways. First, it enables one single operator to control multiple cameras at once, drastically increasing productivity.
Second, the operator can now control the system from any location potentially allowing one operator to remotely work on successive projects no matter how far from each other.
Third, the system is integrable with various cameras, robotic heads, and hardware panels which translates to low implementation costs.
Fourth, thanks to the Presets feature, tedious and repetitive tasks can be automated.
Last but not least, autonomous tracking ensures that the subject is always in the frame. Have we got your interest? Learn more about our offer here:
About Seervision
Seervision is a Swiss-based deep-tech company with a clear focus on the professional video and broadcast industry.
They have combined computer vision, machine learning, model predictive control, and cinematography. The result is their adaptive motion control technology, which won the NAB 2019 Product of the Year Award & the Swiss Technology Award. It transforms any camera into an autonomous camera! The hardware-agnostic technology represents a revolutionary change in studio and sports production and leads to consistently high quality at significantly lower production costs.
Linkedin: https://www.linkedin.com/company/seervision
Instagram: https://www.instagram.com/seervision
