Amazon could develop an app that lets people virtually try on outfits

Image courtesy of Megaflopp

Amazon has filed a patent for an app which could let people try on clothes virtually, using composite images derived from photographs.

The app could be used as a way for shoppers to see what they might look like in clothes they are searching for online, rather than only being presented with an image of a model dressed in the clothes.

Combine tops and trousers from different photos

The patent, which has been filed with the Intellectual Property Office, is for a “browsing interface” that could create a composite image of a person in different outfits by putting together their top and bottom half, based on their existing photos.

It could scour photos a person has saved on social media or on a device such as a smartphone, use the data to learn a person’s dimensions and existing clothes, and then “present hypothetical looks for the user for which no single image exists,” reads the patent.

For example, it could take the bottom half of a photo of a person dressed in shorts, and the top half of another photo of the same person wearing a long-sleeved shirt, and display the two halves together.

“This presents a new way [of browsing] a collection of images to identify new combinations of items to be worn together,” the patent reads.

Scour shopping websites for outfits

It adds that the app could also work with clothes that are “available for purchase” in order to recommend entirely new outfits, which suggests that it could link up with shopping websites.

The app could additionally pair clothing the user already has with suggested new items, to see how they might look together.

“For example, the first portion may show the user wearing [their] favourite red shirt, while the second portion may show an image from a catalogue system of a popular […] pair of white shirts,” the patent reads.

Pick clothes based on weather and social events

Users may be able to buy a whole “look” or style consisting of multiple items through the app.

Controls within the app might include options such as “buy look”, “save look” and “find more like this”, which could suggest similar outfit options. It may also include options to “review” and “rate” items.

The “looks” the app suggests could be tailored to different scenarios, such as seasons of the year or upcoming events in a person’s phone calendar. This could also be based on data it accesses from social media or on a phone.

For example, if someone has meetings coming up in their calendar, images showing “professional wear” may be displayed, while if someone has added a social event such as a night out, “leisure or club wear” might be given preference, the patent says.

It could also be programmed to suggest clothing based on the temperature and weather, as detected by a person’s phone.

App could “radically change” how we shop

The app may have “swipe” controls, which let people change all or part of the image by swiping left or right, but the patent adds that controls could also include “tapping” or “spoken commands through speech recognition”. It may include the option to share images on social media.

Additionally, the app could combine different body parts from various images, with the goal of creating a realistic likeness of the person, using techniques the patent describes as “segmentations and alignment”.

Niel Wilson, chief market analyst at markets.com tells The Telegraph that an app like this could “radically change the way we shop”, suggesting it could have an impact on physical high street stores, and could reduce the number of people who buy multiple sizes online then return some of them.

The patent has been filed and is waiting to undergo a “first examination” according to the IPO, which will then decide if it is granted.

Image courtesy of Amazon and the Intellectual Property Office (IPO)
Design disciplines in this article
Industries in this article
Brands in this article

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.