Collaborative design (Human/AI)

Collaborative design uses generative models in fashion design for fast visualization and to create prototypes. By using deep-learning architecture that learns from images of garments, it generates designs for new garments by modifying attributes such as the colour, shape and texture.

Business benefits

The design process is often long and complicated. Given that fashion is constantly changing, brands have to keep up with the latest trends and predict consumer preferences for the next season. Fortunately, artificial intelligence can help automate certain areas of fashion designers’ work, thereby increasing the efficiency of the process, which in turn brings down costs.

Collaborative design systems can also be used to combine the designers’ expertise with algorithms’ ability to process and analyze large amounts of data and images very quickly so as to identify upcoming trends, thus sparking people’s creativity and optimizing the design process. Working in this way, retailers are given an assured advantage not only in operational terms, but also in terms of speed.

It is not about replacing human procedures, but about augmenting them through technology. Steve Laughlin, General Manager of IBM’s Global Consumer Industry, explained their collaboration with Tommy Hilfiger in an interview with Forbes: “AI can assist design teams by enhancing and reducing overall lead times, and expand their creative discoveries by analyzing 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.”

Algorithms are capable of detecting aspects and patterns that the human mind cannot comprehend or absorb, thus allowing designers to create entirely new design concepts that are sure to delight their target consumers.


Artificial vision
Data mining

Challenges tackled

Cutting down the time to market


Automating and optimizing processes


Increasing sales


Reducing the environmental footprint


Successful cases

The fashion start-up Stitch Fix’s “hybrid designer” draws up a catalogue of their entire inventory and divides up each garment into 30 or 80 characteristics such as colour, length, number of buttons, hem shape, fabric, pattern, sleeve type or collar type. An algorithm then evaluates which of these features are most popular with customers and checks to see if a particular inventory item has several of these popular features. If not, a gap has been found in the market, which means a great sales opportunity. Based on the principle that “more is better”, Stitch Fix assumes that the more of these popular features an item has, the better it will sell. The opportunities identified are then presented to the designers instead of a mood board as the starting point for the creative process.

stitch fix artificial intelligence design

Blouse design generated by the Stitch Fix algorithm from two “parent blouses”. Source: Stitch Fix

Zalando aims to create a streamlined design process that combines the its designers’ invaluable experience with the power of machine learning. Specifically, their team of researchers is currently focussing on finding new ways to use generative models for fast prototyping and visualizations. These models enable machines to learn patterns from the data used to train them in order to generate new similar data. In the case of Zalando, the data used is the images of the items in its inventory. The models trained with these images generate realistic garment designs, but Zalando is now working to keep a check on all aspects of the context for fashion design. Its goal is to monitor clothing attributes such as colour, texture and shape, as well as to work out the effects of these attributes, so that each of these characteristics can be handled independently.

Together with IBM and the FIT (Fashion Institute of Technology), Tommy Hilfiger launched the “Reimagine Retail” project. Using an archive of 600,000 catwalk images, 15,000 brand product images and 100,000 patterns, artificial intelligence was used to automatically generate many designs of new silhouettes, colours, prints and patterns. With this additional source of inspiration, the designers created new, inspiring products that they would not have created on their own. “As a brand, we are always pushing the boundaries of what is possible through innovation and disruption. These young designers truly embody this spirit by showcasing the successful integration of fashion, technology and science,” Tommy Hilfiger’s Brand Director Avery Baker wrote in a blog for IBM.

Annakiki artificial intelligence design

Anna Yang with some models from her “Fashion Flair” collection

Anna Yang, the designer at Anakiki, teamed up with Huawei to launch “Fashion Flair”, their first collection created jointly by man and machine. To do so, artificial intelligence was trained with iconic fashion images along with a specific set of images from the designer’s previous collections. When the images have been processed, the tool has the capacity to create infinite proposals for outfits, presenting endless creative possibilities. All Anna Yang had to do was to select some of those proposals and give them the final touch to turn them into finished garments.

Talk to us

Get in touch with the team to see how we can help you.