Google and Stanford research working separately are forging ahead in the field of image recognition software. Image recognition software is a method by which a computer may identify objects by collecting data and searching a database to identify that object. Google and Stanford using natural language processing and merging it with neural networks have created advanced AI networks capable of accurate product recognition.
This new advance in the software allows it to teach itself to pick out an object from a group with great precision, whereas previous versions could only recognize a single object. After the software recognizes the object, it will write a caption beneath it. Already large retailers are taking advantage of the new advances in image recognition. It’s revolutionizing online shopping, and a user can simply snap a picture with their cell phone to begin a search for that item. The software searches massive databases of images to find a match. When the match is found the information about that item is returned to the user allowing for real-time shopping by the consumer.
Slyce a visual search provider was founded in 2013, and it’s working with six major retailers to bring real-time shopping to their customers. Visual search providers like Slyce work with retailers to provide technology to work with images streamed in by a consumer. Proprietary image technology is integrated into existing retailer applications and databases, and when a consumer scans in an image, it’s streamed to an established database which enables it to match up with the scanned in picture. The matches are exact, and they enable the shopper to proceed and purchase the item on the spot.
Slyce has four retail applications available that can be downloaded to use when shopping with their retail clients. They are all iOS and Android capable, but each provides different services. The application SnipSnap clips coupons and allows the consumer to redeem the coupon right from their device. Pounce searches databases to pick out a product from an uploaded image. Pounce and Craves search fashions databases at retailers like Macy’s and Neiman Marcus and allows real-time shopping. Slyce’s Drivetrain application works on the retail side.
The new research and advances by Google and Stanford in image recognition software is making it possible for companies like Slyce to bring the advances to industry and the public. Soon a consumer will be able to focus on one object out of a field objects and then search the web for the information it needs to buy that product in real-time. It’s a complex process, but no longer out of reach.