The IKEA Place app has gained a Cornell connection.
Using machine learning to understand and recognize objects, materials and style, Sean Bell Ph.D. ’16 and Prof. Kavita Bala, computer science, developed GrokStyle, a company currently developing an application that visually recognizes furniture. As the company works to break into the home furnishing market, GrokStyle has landed a partnership with IKEA, and its point-and-search functionality is being added to the IKEA Place augmented reality app.
According to its website, GrokStyle allows customers shopping for products to use “mobile, AR, web, or in-store” technology to find the items they need simply from images, be they “low quality mobile phone pictures” or “stock photography.”
In other words, a picture of an item snapped from a phone can be used to locate products similar to that item.
In December, Bala and Bell were invited to IKEA’s corporate offices for meetings with company stakeholders, including Michael Valdsgaard, IKEA’s leader of digital transformation.
“[Valdsgaard] immediately recognized the relevance and impact [GrokStyle] could have on the consumer experience in IKEA’s Augmented Reality app, IKEA Place,” Bell told The Sun in an email.
According to a GrokStyle press release, Valdsgaard was “impressed.”
“After one meeting and demo with GrokStyle, I was convinced we had to do this now,” Valdsgaard said, according to the press release.
After a demonstration of GrokStyle’s technology, a partnership was created and in just three months, the app’s technology was integrated into IKEA’s system. IKEA Place 2.0 launched globally through a “collaborative effort” of Inter IKEA, GrokStyle and two other companies.
“This partnership puts visual search into the hands of millions of IKEA customers across the world and so dramatically increases consumer awareness of visual search as a valuable tool when shopping for home furnishings and decor,” Bell said.
Augmented reality is currently a developing area in home design, furnishing and decor, as industries explore ways to help consumers visualize their space before making purchases. IKEA marks “the first of many retail collaborations where we reimagine and enrich the customer experience,” Bell and Bala said in a statement.
“A number of pilots are currently underway to incorporate visual search into the apps and web solutions of leading retailers,” Bala told The Sun in an email. “We will also continue to collaborate with IKEA to expand the reach and impact of visual search.”
GrokStyle hopes to continue advancing research in “building visually-based artificial intelligence solutions for retail,” including through the development of a product and style recommendation engine, creating prediction tools on what customers are looking for based on visual search and expanding the application of GrokStyle’s algorithm to different product categories.
Bala explained how visual search technology can help consumers overcome issues with using words to describe a desired product.
“Our frustration with language gets in the way of our exploration of design decisions, and makes people dread having to shop for anything,” Bala said. “Visual search is giving people confidence because it is far easier to search for a chair that looks like one at the coffee shop, or goes with a particular coffee table, for example, than to put the words together that will indicate those ideas.”