Artificial intelligence at the service of trend detection

Posted by The Fashion Heaven on Tuesday, September 25, 2018

On February 13, 2018, the Journal du Luxe organized a conference on the Opportunities of Artificial Intelligence for Luxury. Although not exhaustive (because to access all the content and participate in the debate, we invite you to participate in our events – You can subscribe to the newsletter of the Journal du Luxe to be sure to receive the invitations), we welcome you offer access to 4 exclusive articles containing 4 major topics raised on this occasion. After dealing with the subject of artificial intelligence and its relationship to creativity and emotional data and their application in the luxury business, we are now interested in AI at the service of detecting trends. Back to the intervention of Tony Pinville, co-founder of Heuritech on “How does I.A drive branding? ”

Season after season, trends evolve, fade and reappear. Some leading figures in the sector, both renowned creators but also influential personalities, have long launched fashions. Today, influencers (especially on instagram) launch the trends and they are thousands around the world! Social networks allow communities to express themselves, to converse and especially to assert their tastes.

Tony Pinville takes the example of the return of the banana bags on the front of the stage. Worn by stars like Bella Hadid and posted on social networks, they are once again gaining popularity in clothing styles. Trends therefore continue to evolve, but they are no longer the consequence of choosing a handful of people. By generating debate and interacting, society as a whole is at the origin of this collective problematic.

Tony Pinville is co-founder of Heuritech, a start-up specializing in deep learning and awarded the LVMH Innovation Award in 2017. Heuritech’s job is to identify the latest trends in clothing style, colors, types products, etc. to then transmit them to the marks. By using artificial intelligence, Heuritech will be able to detect a product among thousands on social networks, differentiating it from the same typologies of products offered by different brands. In practical terms, the tool will make the difference between a Supreme Banana bag and an Alexander Wang Banana bag.


The results are triple because they allow: – an in-depth study of trends in space and time, – segmentation of customers by trend and geographical area and finally – an analysis of the product performance in terms of online radiation.

What is the interest in branding? “We help product development, better understand the behavior of our customers and also assist them in merchandising. ”

Tony Pinville – Co-founder of Heuritech

Where the analysis becomes even more interesting, it is by crossing the data to identify the environment in which the product is located: is it photographed day or night, in what context, with what other types (and brands) of is clothing or accessories associated with the photo in question?

Thus, the use of AI in detecting trends (globally) also allows brands to identify emerging trends within targeted people profiles to enrich product development and better understand their customer base.

Finally, other tools, such as Data & Data, make it possible to identify products put online on the gray market, for example.

The concrete uses of AI in the detection of trends (of all types) are therefore broad and its association with other technologies suggests many uses in the near future.