For those who haven’t heard of the most frequent application of image recognition is the commercial visual search engine called “Slyce.” Slyce is actually a larger player in the image recognition app category as they recently acquired SnipSnap, the mobile coupon reader that had 4 million mobile users in its own right.
Slyce is an image search engine that is using image recognition, which is still growing by leaps and bounds as a basis for search. You know it’s on the forefront when Zuckerberg outlined his artificial intelligence plans for Facebook at the annual developer’s conference this year. Zuckerberg showed a sample of image recognition that allows the blind to hear explanations aloud of images on the screen. He was quoted as having said about his AI plans, “we will build systems that are better than people at perception.”
One of the larger adopters of the Slyce app came from over 2,200 participating movie theaters that enabled their moviegoers to read pictures of interactive ads in the lobbies of movies. You can get information such as where to buy the types of things as worn by the stars in the movies your seeing.
This is just but one application of visual search intelligence, but conceivably most businesses could find an image appliacation. Image recognition is new but really expanding due to the climate of open source development. So much so that for the third year in a row there will be a LDV Vision Summit, an annual conference to cover visual tech, VR, cameras, medical imaging, and content analysis.
The visual search field is set to be one that really has a monumental impact on business – get more at Slyce. This is due to a few different factors. First, is the open data available as training material. Another factor is due to deep learning, a machine learning techniques patterned in a loose way the way a on the human brain.
Deep learning aids in computers learning what is in pictures faster and more accurately. Two particular databases have been of particular importance, Pascal VOC and ImageNet. These free database collections have millions of images tagged with keywords and descriptors inputted for every photo and ready for use – Slyce – Visual Search, Image & Product Recognition
Other factors include having open source software frameworks and libraries to use as building blocks for the technology’s development. Also the vast array of hosted API’s available today will be a great help in furthering the research. Most tech training grounds such as Stanford, Cal Berkley, and Princeton have research on going in the field of image recognition. With all the positives it has to offer, one has to think about the impact to the business world in an exponential way.
Get full details at: http://slyce.it and http://techcrunch.com/2016/04/30/why-image-recognition-is-about-to-transform-business/