The Killer App that kills Computer Vision

What is the shelf life of your computer vision application?

There is a plethora of ideas for computer vision applications and plenty of very smart folks are working towards finding the killer application(s) using computer vision technology. However, there might come a day when an alternative approach would replace computer vision as the more ideal implementation. This alternative approach might be better in performance, easier to implement, simpler to install, and most importantly, more cost effective.

On the one hand, it is likely that the computer vision application itself is replacing a worn out approach with better, safer, and more intuitive approach to the problem. For example, video camera operators for security and surveillance are now being replaced with smart camera technology that are always on, ever vigilant, and continually unbiased. Hours of manual labor spend to analyze live or recorded videos are being replaced with automated means. So it is not surprising that the computer vision application itself is one day replaced. Such is the fact of life.

On the other hand, it is another thing if computer vision applications are replaced before they become commonplace. That is, before computer vision has its debut, it might be replaced already. Consider for example, brain interfaces that are currently touted to read thoughts. Sensors surgically implanted in the brain or mounted on portable devices can detect brainwaves to enable a "brain-computer interface" to control machines by merely suggesting a thought. Brain-computer interfaces have applications ranging from gaming to twitter. They are sought after by the physically challenged (as well as those too lazy to type on a keyboard). They are one step ahead of camera-based gesture recognition system as they do not need the use of a camera.

The point of this article is not to suggest that brain-computer interface technology is better than computer vision, or vice-versa. The intent is to suggest that computer vision scientist ought to move a little faster, or else other technologies might creep up on them. Venture capitalists invest in new technologies as soon as they can identify some competitive advantage that would result from their investment. Funding would be difficult when venture capitalists look at your computer vision ideas and suggest that there are alternative technologies (i.e. more lucrative approaches).