Text search boxes are a staple of the internet. It is how we connect to places on the web. But recent advances in image recognition software may make s big difference in this area. Websites like the Pinterest social network and the online shoe store Shoes.com are testing new methods of performing searches using just images.
A technique called deep learning has made the changes possible. It enables software to match humans when it comes the act of image and product recognition. The new technology is spearheading Google’s image search and photo recognition site it recently launched.
Pinterest’s new tool allows you to draw a box around an image on their website in order to search for similar items from an index of over a billion stored images. A example would be to box a coffee maker seen in a photo of a kitchen. The visual search produces several similar items, including an exact match.
Some search results on Pinterest produce images with buying buttons. The feature is also included in their mobile app. Pinterest believes visual search can become indispensable because of deep learning’s accuracy. Still, the software is not perfect. Experts think that only after the masses have tried visual search will we know how much of an impact it will have on online searches.
Shoe.com is using the deep learning advances in another manner. You click on a visual filter button and you get 12 styles of boots from a catalog of 7,000. When you pick one closest to what you desire you get a selection of another 11 you might like. Shoe.com says the new feature increases sales, and they might expand it beyond boots.
Slyce is a leading visual search recognition company. It is headquartered in Toronto Canada. It was launched in 2013. Slyce develops visual search recognition software based around mobile smart phones.
Large retailers use Slyce’s software to connect its online stores with mobile phone users who take pictures of products that can be searched out and purchased through Slyce’s apps.