Visual Search Startups Competing for a Huge New eCommerce Market

According to an article published recently on Live Mint, startup companies are working on bringing visual search engine technology to the ecommerce sector.

Customers who want to buy a certain product could take a picture of it with their smartphone, then ask the search engine to find products like it, based on their appearance.

Live Mint gives the specific example of Anupama Pasumarthy, a 22-year-old graduate tech-savvy student. She shops online, but not for wearing apparel, because she’s always disappointed by the selection displayed to her. The clothes were too big and the shoes too small. Therefore, to dress herself, she’s returned to brick and mortar stores.

But perhaps she won’t have to resort to that for long. Many new companies are attempting to use Artificial Intelligence (AI) to analyze image files to understand what they display. Some of them are Stylumia Intelligence Technology, iLenze, Mad Street Den, Slyce, Visenze and SnapShopr.

Slyce was launched in 2012 by Cameron Chell and Erika Racicot. It has its headquarters in Toronto, and is publicly listed on the TSX Venture Exchange. Slyce uses its Universal Scanner to scan or take pictures. It can handle digital images of objects, QR codes, coupons and bar codes.

For a two-dimensional image, it will create an exact match. With a three-dimensional image, it will search a retailer’s catalog for the most similar image. 3D images are analyzed for attributes which the retailer defines (dozens available). Depending on the product, these may include size, color, style, pattern and unique features.

Amazon has implemented its Amazon Flow mobile application, but Flow has limitations. When someone snaps a shot of the product they wish to find on Amazon, it must be in its original packaging because Amazon’s product relies on identifying the company logo, the OCR bar code and other such cues. That makes it useful for comparison shopping when someone is in an offline retail establishment. Once a product is taken home, it’s taken out of the packaging.

Slyce does not have this limitation because it analyzes and uses image recognition on how the product itself appears. A woman could admire a friend’s watch, take a picture and then order her own watch, just using her smartphone.

Visenze’s technology is being used by Flipkart, the Indian ecommerce giant. Stylumia is based only on fashion. It analyzes any content that may influence fashion, including TV shows and Bollywood movies, and extracts fashion trends from that data, and then makes its recommendations to shoppers.