GoogleAds - Half Banner


The Diversification of the Computer Vision Killer Application



You can participate on a currently running public opinion poll on the next great killer application in computer vision. But which application is the real killer application? If you were to ask the community, you would find a wide range of answers ranging from camera phones to automotive, from robotics to surveillance. Perhaps the answer to the question is all of the above, or perhaps none of the above.

Over the years, I have spoken about this and have certainly heard a variety of answers. It is not surprising that there is no clear response. Computer vision is widely applicable in many imaging applications, and everyone has an answer to what constitutes the killer app. The research field incorporates elements of machine learning, artificial intelligence, signal processing, optics/imaging, and mathematics. And consequently, computer vision is intricately weaved into other fields such as robotics, visual computing, and human-computer-interaction.

It is the diversification that is perhaps making it difficult for the general public to get excited about computer vision. It is natural for us to associate with technology that we are accustomed with. For new technology, however, it is much more difficult to make the connection for its capabilities. Very often, I have noticed that it is the other fields incorporating elements of computer vision that rallies the general masses. For a case in point, look at the gaming arena. Camera based gesture recognition are now being introduced in game consoles. The smart camera is not the marketing focus, but rather an enabling technology for better user experience. The casual gamer only needs to know that it works, and not how it works.

In the long-term though, I believe the diversification is a blessing for computer vision. In a field where the end goal is still being defined, it is perhaps useful that computer vision algorithms are easily integrated in many applications. For the active researcher, it is nice to know that your career is not about to hit a road block akin to a writers block. We can not afford to focus on any single application when there is no clear winner. Instead, we must progress onward and continue to apply our knowledge as the application dictates.

Your rating: None

The poll on killer computer vision applications mentioned in this article has ended. Results are available here.