Audio version of the article
Scentmatic pushes the perfume industry to the next level.
In recent news, Japan-based Scentmatic aims to shift the perfume industry’s paradigm by collaborating with Biotope to offer its AI system to an exclusive number of Nose Shop Branches, a boutique of beauty products located in the nation.
Precisely aiming to offer its customers new ways to experience scents and aroma, Scentmatic, with Kunitake Saso’s aid, who leads Biotope aspires to achieve the said goal with the assistance of AI. From this ambition comes Kaorium, which is an AI-powered system used to identify the scents which define a perfume. In other words, the AI aims to verbalise the flavours, aromas, and essence of the perfume it analyses.
The reason behind Scentmatic, which is led by Shuji Kurisu, being drawn towards this exercise is the intention to utilise AI to translate scent and fragrance related products. This is to build a library of verbal profiles to define each aroma.
“The world is filled with different scents. But they are often not sensed properly, and thus the chance of fully experiencing them is often overlooked.” explains the company’s vision to shed further light on their motivation of being part of this demonstration with the power of AI.
How does it work?
To initiate AI-powered Kaorium, the user has to choose from 20 different bottles filled with unique aromas dispensed from the boutique. After selecting from the collection of fragrances, the customer is allowed to select words that are felt when the patron is inhaling the said aroma.
The outcome of this interactive exercise is to allow the customer to have the freedom to attain a favourite scent that the AI algorithm can clearly describe. The result can be said to have impressive accuracy as the software evaluates up to 5000 subjective data, which has been harnessed from the various patrons of the Nose Shop.
With such data backing the AI, the clarity of the scent which the customer has felt can be reinforced. Utilising data and AI for such a unique experience, the mission of Kaorium is to create a new axis in selecting fragrances.
Delving deeper into the workings of Kaorium, the database is a place of exchange where the scents are connected to words that best define them. With the user interacting with the AI, an exchange of images and words allows the patron to best define the fragrance they are seeking. This conversation is repeated over and over again allows the user to refine one’s favourite scent. The output is the AI recommending a mixture of scents to create an aroma personalised to one’s flavour.
This is not the first time that a perfume company has been seen harnessing AI’s power to produce better fragrances. To witness the genesis of AI being deployed to permute a fragrance, we can look at the collaboration between IBM and Symrise in 2018, which created the AI named Philyra. This software can learn new formulas of scents, identify raw materials, display each fragrance’s historical performance, and define market trends in the perfume industry. The AI employs a variety of complicated machine learning algorithms to identify trends and perfume permutations by fleetingly looking through various combinations.