Look around you and try to put a name to the things you see. Chairs, tables, floors, cups, plates, they’re all easy to recognize. You can probably even recognize the brands of many of those items when you see them. It’s a fairly easy task for people. But it’s a shockingly difficult task for computers. Many people aren’t aware of how many decades of serious research have gone into the seemingly simple task of computers being able to recognize a chair when they see it. This process, known as image recognition, has been one of the most complex problems within the field of artificial intelligence.
The past few years have seen some amazing advances though. What used to be a distant dream is now a reality. The right software, created by the right people, can finally recognize and define the objects within a picture. Of course that’s impressive on a technical level. But people might wonder what that means on a practical one. What can image recognition actually do? By itself it really is just an academic accomplishment. But some companies are working on ways to integrate this new technology with existing platforms. One of the most powerful comes about through the use of mobile phones and image recognition.
There’s been two noteworthy changes in smartphones which have made this possible. The first is that the quality of cameras has gone up tremendously. The cameras in smartphones were often more of an afterthought. Today the cameras in smartphones rival professional quality equipment. This ensures that the image recognition software has high quality images to work with. The other improvement happened with networks. The software on the phone isn’t even close to powerful enough to run high end image recognition software. So the image needs to be offloaded onto higher powered equipment. Higher quality data connections make this trivial to accomplish. It just takes a few seconds to snap a high quality picture and send it off to a more powerful remote server for processing. Of course that requires a company to do the processing.
The current industry leader in image recognition in phones is Slyce. They make it easy to work image recognition into an app. For example, one of their services involves product recognition. They can enhance or even build apps which will notice a particular item, search internal databases, and lead potential customers to a location where it can be purchased. But their “snap to buy” service is only the beginning of what they can do. They have specially tailored services to match almost any image recognition need.