Home Featured AI: the Antidote to Fashion’s Fear of Change

AI: the Antidote to Fashion’s Fear of Change

0

In her first exclusive for 2018, resident Expert Dr Evridiki Papahristou explores Artificial Intelligence, in direct relation to our industry. Evridiki is a devoted fashion engineer with a research focus in the effective integration of 3D virtual prototype in the apparel industry, and sits on our Expert panel in implementing and adopting 3D.

When I think of AI, the first sci-fi movie that comes to mind is Spielberg’s Minority Report with Tom Cruise and the Pre-Congs predicting the future. In 2002, the ideas of highly targeted digital advertising, big data, analytics, and the IoT were relegated to science fiction. In the year 2054, commuters are bombarded with personalised ads in train stations. They are identified with strategically paced retinal scanners. The protagonist, Tom Cruise, is greeted by a virtual sales person in a Gap store. The Minority Report scenes provide a glimpse into what’s possible when context includes deeper knowledge about the individual than what traditional digital channels can deliver.

New Era / New Consumer Needs / Technology Revolution

Now back to 2018. As a fashion engineer and researcher I constantly try to find ways to apply technologies from disciplinary scientific fields, proposing new business models in the fashion industry. It is only recently that I wondered: was Spielberg right? Are these scenes, precisely as depicted, a likely future state?

What is real is that visual search for shopping feels like the beginning of an industry trend. What is also more than real is the fact that consumers demand the latest style in double quick time and, because of the rate at which trends change, speed of delivery has become quintessential to retailers. The human mind though, cannot comprehend or retain thousands of images, not to mention such amount of information data.

Fear of Change (the Virus)

Those human minds, apart from being traditional and old-fashioned, are stuck in the processes of the past, creating an oxymoron for the forward and open-minded industry of fashion. These old systems dig their heels in, blocking and frustrating access to renewal.

The Current Situation with A.I

By 2019, more than one fifth of the American population will use Siri, Cortana or another virtual assistant at least once a month. “Consumers are becoming increasingly comfortable with the technology, which is driving engagement”.

We now see millennials following even computer-generated influencers like Miquela Sousa or better known by her Instagram handle @lilmiquela. Miquela appears in her Insta account wearing Chanel, Prada and Supreme and she even works with fashion magicians and advocates for social change. What is different? She is a virtual avatar!

If you think that no one has thought of implementing AI technologies in creating new fashion styles then you have missed the 1st season of the AI & Fashion series. First Amazon, with the Lab126, a research centre based in San Francisco, has developed an algorithm that learns about a particular style of fashion from images, and can then generate new items in similar styles from scratch—essentially, a simple AI fashion designer.

Google, in partnership with Zalando, with Project Muze, first trained a neural network to understand colours, textures, style preferences and other “aesthetic parameters” derived from Google’s Fashion Trends Report as well as design and trend data sourced by Zalando. Then, the project used an algorithm to create designs based on users’ interests and alignment with the style preferences recognized by the network.

Source: http://www.instagram.com/lilmiquela/

Tommy Hilfinger has teamed up with IBM and the Fashion Institute of Technology (FIT), Infor Design and Tech Lab, on a project called Reimagine Retail, using AI to predict which items are going to be in style in the coming months (and even years).

StitchFix, an online personal shopping service, picks clothes for each customer from a sea of inventory with the help of a set of algorithms. These algorithms, which transformed a small startup company into a public one in just seven years, help recommend items within each customer’s price, size and style preferences. The technology behind this robot-assisted styling tool is now being used in the creation of new designs – based on combinations of consumers’ favourite colours, patterns, and textiles.

On the retail side, AI is already blurring the line between online and offline. Alibaba developed FashionAI, a personal stylist. The technology of deep learning has taught FashionAI to recognise hundreds of millions of items of clothing as well as the tastes of designers and fashion aficionados on Alibaba’s shopping sites, and then makes outfit recommendations to clients, based on previous activity and the store’s inventory.

Source: http://algorithms-tour.stitchfix.com/

Researchers from UC San Diego and Adobe have outlined a way for AI to learn an individual’s style and create customised computer-generated images of new items that fit that style. This represents a first step towards building systems that go beyond recommending existing items from a product corpus, but which can be used to suggest styles and aid the design of new products. They believe that this model opens up a promising line of work in using recommender systems for design.

Artificial Intelligence (the Antidote)

The problem

The Fashion and Textile industry is unconsciously biased towards applying AI technology solutions, especially in the product development process. In a Google presentation on bias that I attended recently, they estimate that at any point in time our brain is processing some 11 million bits of information and we can only consciously process 40 bits, which makes us 99.999996% unconscious. According to this statement, imagine the consequences if we didn’t have any tiny bit of bias. The positive part is that there are methods to overcome our unconscious bias and yet another challenge for our industry.

Source: Google re:Work

The Solution – A.I (the Antidote to the Fear of Change)

In Minority Report, a neural network could identify criminals. In the real world, could it identify the next top fashion trend? Could it even help to create new styles that could target best the customer’s preferences?

Making trend prediction more accessible and getting products to market faster means fewer lost sales. One of the big questions that rise up around the use of AI in design, decision making and product development, is whether it spells the death of creativity. Whether it will ever develop the wits to be truly creative.

All this poses a wider question. Should it matter to brands, or the consumer, if the inspiration and ideas are computer-generated? If AI algorithms predict style trends rather than a “real” fashion designer?

Change by Differentiation

Companies employ various means of differentiation in order to gain a competitive advantage in the market. If a fashion company is going to differentiate by data science it cannot do it as usual, with the same old processes, structures, roles etc. The traditional organisation needs to be changed. Oh, and here is the same old word: “change”.

According to Steve Laughlin, General Manager of IBM Global Consumer Industries, AI can assist design teams by enhancing and reducing overall lead times, and expand their creative discovery by analysing and remembering insights from thousands of images and videos using computer vision. These designers can also more easily find how they can integrate trending colours, key patterns, and styles. It’s about reducing a time-consuming, resource intensive, manual process – or blowing up that research element by providing access to much wider sources than ever before, he commented.

Clearly, much more R&D is needed before brands rely on AI-only designers. But artificial intelligence is already helping brands create and iterate their designs more quickly. Soft skills like creating and designing are going to be more important in the future. However, a selective use and interpretation of data to supplement the human creative insight (but not replacing it) and help make data powered decision, may be the only choice in the competitive world of fashion.

If we want to kill the virus of fear then we must sacrifice something. “Without sacrifice, there can be no victory” …but that’s from another sci-fi film…

Evridiki Papahristou Dr. Evridiki Papahristou is a devoted fashion engineer with a research focus in the effective integration of 3D virtual prototype in the apparel industry. She began her studies at the University of Kent in European Fashion/Product development but after working as a fashion designer in a design studio in Milan she continued her studies with a Master’s degree researching new technologies in apparel. Working in parallel, as a CAD systems trainer in the industry as well as a lecturer for more than 15 years, she has trained many of today’s executives in digital fashion product development.Her recent Ph.D is a unique blend of fashion and engineering, trying to explore new ways of implementing and adopting new technologies like 3D and PLM. She loves research and she strongly believes that knowledge should be shared diffusively without restriction.