Amazon is ramping up its efforts to become a fashion giant and in typical Amazon fashion, is taking an algorithmic approach. According to a recent article in MIT’s Technology Review, Amazon researchers have created machine learning software that can determine whether a particular look or outfit is stylish or not. In addition to teaching a computer what it means to have style, Amazon’s research and development center Lab126, has already developed an algorithm that can generate new fashion items after reviewing images of similar styles:
This work uses a cutting-edge tool called a generative adversarial network, or GAN. It consists of two deep neural networks operating in tandem to learn efficiently from raw data. The GAN internalizes the properties of a particular style simply by looking at lots of examples, and it can then apply that style to an existing item of clothing.
While Amazon’s new tool cannot create a new fashion trend, it can design new items based on current ones. This could be particularly valuable considering the rise of fast fashion. Successful fashion retailers, like H&M and Forever21, release new styles daily. Using artificial intelligence to expedite the creative process could reduce the time it takes to design new items and keep Amazon one step ahead of the competition.
The Echo Look is Key to Amazon’s Fashion Empire
In addition to these machine learning tools, Amazon also has another powerful tool at its disposal: the Echo Look, Amazon’s virtual style consultant. Using machine learning and real human experts, its Style Check feature helps users pick stylish outfits as well as suggesting other items that would work with their current wardrobe.
As Voicebot readers know, a key ingredient to creating an effective virtual assistant is lots of data. Using the data from their customer’s purchase and search habits for fashion items on its site is helpful, but they cannot effectively see how they are wearing the items. But with the Echo Look, customers are willingly giving Amazon access to their closet and whether they know it or not, their wallets. This kind of information is priceless. Using this data combined with tools like the one developed by Lab126, Amazon could literally design new items based on their actual customer’s wardrobe and current fashion trends. If a customer regularly likes what Amazon suggests, you can bet they’ll keep coming back for more.
In fact, it’s science. Research has shown that the pleasure center of the brain lights up during the shopping process for two reasons: getting a good bargain and getting what you want. Amazon is working to ensure it achieves both.