Amazon is developing a fashion app that will “revolutionise” clothes shopping by allowing customers to try on outfits with a ‘virtual mannequin’ created from social media photographs.
In a development experts fear will threaten the High Street, the American online giant has lodged a UK patent to scour “selfies” and online calendars to predict clothes a person may like or need.
It would “data mine” pictures saved on a phone or computer to produce an Augmented Reality image of the customer actually wearing clothes on sale, potentially eliminating the need to visit a shop’s changing room.
The app would also analyse photographs and appointments to establish a person’s job, the climate they live in, how they spend free time and suggest outfits for upcoming events or accessories for other things they own.
The patent application, seen by the Sunday Telegraph, says customers could swipe to like or dislike garments their mannequin would be shown wearing.
It explains “the number of images digitally captured continues to increase”, in part due to “arms-length … selfies”.
Credit:
Michael Heim/ EyeEm
They could “search my look”, “randomise” outfits, and “find more like this”.
After granting access to personal data, the app acts as a personal shopper, showing them “wearing an outfit for which no image actually exists”.
Photographs of a customer wearing a variety of styles – such as T-shirt and shorts or long-sleeved shirt and trousers – would be harvested to “provide a realistic view” of them wearing similar clothes for sale.
Credit:
Amazon
The algorithm draws on personal information “sensed or otherwise provided by the access device”, to “determine the temperature or season” in photographs to better predict what they need.
“If the access device determined the temperature was hot, it may be more appropriate to display warm weather clothing rather than skiing clothes,” the patent application at the Intellectual Property Office says.
“If a [calendar] event was entitled ‘business meeting with James’ the images showing professional wear may be selected. If it was entitled ‘dancing with friends’, leisure or club-wear may be given preference during the selection process.”
Credit:
Jeff Chiu/AP
The app could even scour devices belonging to a “friend or family member” to allow the “user to find an item of clothing to borrow”.
Mark Howell, creative director at strategic design agency Play Retail, said an “interactive personal mannequin” could “revolutionise” clothes shopping.
“The potential amount of data that this technology will be able to generate for Amazon is staggering,” he said.
“By having access to a customer’s photos, the retailer can see what they wear, what they drive and even inside their home to better predict what products an individual is most likely to want to buy.
“It would analyse pictures to determine segments of the body to create a composite image.
“Crucially, it would enable you to ‘try-on’ new outfits digitally.”
Credit:
Amazon
The app would also enable people to share recommendations, wish lists and favourite outfits via social media.
Neil Wilson, chief market analyst at Markets.com, said such an app could “radically change the way we shop” potentially “accelerating the shift” away from the High Street.
It could reduce the number of so-called “serial returners” who order numerous styles and sizes, before returning all but one outfit.
Credit:
Amazon
Andrew Busby, founder of Retail Reflections, said many consumers would welcome the move, despite it posing a “real threat” to some stores.
He added that Amazon had begun “disrupting” the fashion industry sector, with “more traditional clothes retailers” facing “big repercussions” from advances in technology.
“We are seeing Darwin’s survival of the fittest being played on the High Street. The strong will survive, in part because years ago they embraced the internet,” he said.
Amazon refused to comment on the patent, which is in its early stages.