* Video modification technology from Xunavision
The adage that the “camera never lies” is now well-known to be pure fiction and indeed anyone who knows anything about the art of photography will be fully aware that the camera has almost never told the truth!
At the heart of this untruth is the very profound need to tell a story, convey a message, or, as is very often the case, sell something – a product, a service, or an idea. Whatever the reason, humans will always be perverting images to make them do what they want.
Now we’re all familiar with what can be done with image manipulation tools such as Adobe Photoshop, but most will only have a passing familiarity with tools such as Adobe After Effects for the simple reason that manipulating video is complex and expensive.
Let’s say you shoot a video that you’re going to put on YouTube and then a large company, say, Saks, offers you a gazillion dollars to feature their logo on a bag in the video instead of, say, Neiman Marcus’s.
The solution that would occur to most people would be to re-shoot the scene as the normal financial and time cost of modifying video is enormous. Given a gazillion dollars it would have been cheaper to just start from scratch.
Well, technology marches on and a group of researchers from Stanford University have created a technology called Zunavision. They have created a service that allows anyone to upload a video to the Zunavision Web site and then embeds either static images or videos within that content. The embedding is done so that the included content’s orientation relative to the surface it is applied to is maintained even as the viewpoint and lighting changes.
What this means in our gazillion dollar brand replacement example is that is the area containing the logo on the aforementioned bag would be selected and new logo inserted. Then, as the camera angle changes, the embedded logo will be morphed to conform to the surface as the surface changes size and angle. Even more impressively, if a shadow passes across the bag or the lighting level changes the embedded logo will be shadowed, dimmed, or lightened as appropriate.
The Zunavision technology is still under development and the developers currently only allow limited testing. In the current implementation the video output if currently only suitable for Web video. Even so, the results are beyond impressive.
The developers note that among the applications is the possibility of users modifying their own videos to include advertising for revenue generating purposes – the technology is that easy to use that pretty much anyone can do it.
This is a technology to watch as it potentially not only has a powerful place in professional video but could change the nature of user generated content in YouTube profoundly – when every video contains a message or an advert obviously displayed or subtly hidden and when every frame is a potential revenue generator we’ll have even less reason to believe that the (video) camera never lies.
