Category Archives: Shopping Apps

How Algorithms Are Replacing Text Searches with Pictures

Published / by FitGuru

Google famously started Google Images after the massive response to Jennifer Lopez’s green Versace dress, worn to the 2000 Grammy Awards.

Google being Google, they quickly applied algorithms to the behind-the-scenes data available in each image file processed. Soon, those algorithms expanded their scope to account for the visuals themselves. Before long, “search by image” became a reliable way to source original pictures, as well as any images related in both composition and subject.

However, now that the technology is available, what can be done with it? Visual search on slyce.it recognition systems are desired by everyone from Hollywood effects studios to the military. In between, though, are a range of industries that can benefit from the technology. With the amount of shopping being done on the internet, connecting what people see to where they can buy it is developing quickly.

One company making a go of it is Slyce (slyce.it). Their app(s) utilize a variety of approaches for visual search, while attempting to encompass every option available to consumers for identifying a product. For example, QR codes are everywhere, yet have struggled to justify their existence on their own. By incorporating them into a broader image-identification engine, Slyce is establishing a multi-pronged approach to product identification.

The more impressive component of this, of course, is the ability for users to take any image, highlight the element they want identified, and then find a range of results from the massive superstore that is the internet. It could be a picture found in a magazine at the doctor’s office; a screenshot from a television show; or even a friend’s picture on Facebook.

The benefit of this technology for commerce is flexibility. Slyce offers a Neiman Marcus-branded app that focuses on the products available in their own lines. With a larger retailer like this, searches of popular styles can find exact or near matches without compromising on what the consumer wants.

Google has already shown what a wide-open system is capable of. Specialization of these systems allow their creators to manage the content in a way that provides the best deals to consumers without concerns over scams and out-of-date results.

Of course, specialization is what the internet does best. It’s hard to imagine a more design/product-focused library of pictures than Pinterest. The news that they’ve incorporated their own visual search is simply another sign that text searches have gained a powerful new friend.

Slyce a Toronto Startup

Published / by FitGuru / 1 Comment on Slyce a Toronto Startup

Slyce is a new Toronto startup. According to the chief digital officer, Mark Elfenbein, the idea of the company is to give customers the ability to take a picture of any product in the world, using product recognition, and all users to purchase the product. However, unlike the popular app Shazaam that helps users find music and send it to their iTunes store, the Slyce app will not be seen my users. Instead, the new technology will be offered in conjunction with bigger brands’ websites.

The way the technology works is by determining what the item is at first by looking at it. Then it will determine what the properties of the item are. For example, it will first make a determination of whether the product is a towel, plate, or shirt. After that has been narrowed down, the technology will analyze the image according to Slyce’s schema that was programmed for each item.

When it comes to a shirt, the technology might analyze the pattern, color, location of pockets, and more. By breaking each product down like this, Slyce is able to provide users exact matches and other objects that are similar. However, the technology must first know what it is supposed to be looking for.

Mr. Elfenbein recognizes that every they venture into a new market there will always be a certain amount of work that needs to be done to build a technology before it is released. Therefore, the company must build models for each item. After it has been built, it can be reused over and over again.

When it is all said and done, the new technology is being built into retailers’ apps and online stores. Slyce is not stopping with helping users match pictures with the actual product.

The data for the retailer is almost as equally important as whether the user buys the item or not. The technology can be used to create lists for wedding registries and other special events.

While Slyce is a Toronto based company, it is starting to spread throughout North America as well. Mr. Elfenbein also recognizes that the company is growing very quickly. He believes that the new technology has the ability to revolutionize how users interact with the brands they are interested in.