Interview: We’ll tell you more about the winner of the “The Cosmetic Victories” competition.

Latest News

PhD student Matej HLADIS from the Université Côte d’Azur wins the grand prize at the “The Cosmetic Victories” competition thanks to his innovative artificial intelligence project capable of predicting scents.

After completing his studies in applied mathematics in Slovakia, Matej HLADIS pursued his passion for research by starting a PhD at the Institut de Chimie de Nice, affiliated with the Université Côte d’Azur. His academic journey led him to delve into the fascinating field of olfaction.

Despite the initial challenges requiring an in-depth understanding of chemistry, biology, and neuroscience to decipher the complex mechanisms of the sense of smell, his strong foundation in mathematics proved to be a valuable asset. Thanks to his skills, Matej quickly grasped the intricacies of deep learning and artificial intelligence, thus propelling his project to new heights.

What is the “The Cosmetic Victories” competition about?

The goal of the “Cosmetic Victories” is to offer researchers and startups the opportunity to present their innovations to the main players in the cosmetics and perfumery industries. France is one of the global leaders in this sector, and to remain so, the Cosmetic Valley, which organizes the event, hosts several events to stimulate collaboration and innovation in the field.”

Can you tell us a bit more about your award-winning project at the competition?

The goal of our innovation is to introduce AI into the perfume industry. In our work, we have developed a bio-inspired AI model capable of predicting the scent of any odorant molecule, based on our understanding of the human olfactory system. To put this problem into perspective, there could be more than 40 billion molecules that have a scent, and so far, we have explored only a tiny fraction of this colossal number. Thus, this approach can open up entirely new types of scents that we have never encountered before. Additionally, it can help perfumers and experts find more sustainable and safer alternatives to the odorant molecules already used in the industry.

How did you get the inspiration to develop this project and this AI?

Our research group’s expertise focuses on studying olfactory receptors, which form the basis of the combinatorial code of olfaction. Humans have about 400 different types of olfactory receptors in the nose, and our sense of smell operates through a complex coding mechanism, where each odorant molecule activates some of these receptors, and based on the combination of activated receptors, our brain receives a signal, processes it, and ultimately, we have a sensation that we call a scent. The core of my work is to use the latest advances in deep learning and AI to predict this combinatorial code for any odorant molecule. Recently, we have achieved very encouraging results in this task, and a natural next step was to use our knowledge of the olfaction mechanism to finally predict the scent of an arbitrary molecule.

What impact do you think your innovation will have on the perfume and cosmetics industry?

Currently, we know and use only a small fraction of all possible odorant molecules, and it is very difficult to explore new scents because currently, we would need to first extract or synthesize the molecules and then smell them. As there are millions of possible odorant molecules, this is not practically feasible. Our innovation can achieve this in just a few days. Thus, perfumers would have a tool to rapidly explore entirely new olfactory territories that remain uncharted. Furthermore, a second challenge is the formulation of molecule mixtures. Our sense of smell is very complex, and mixing two molecules can produce a scent very different from that of its components. Our innovation opens a new way to tackle this problem, bringing us closer to computer-assisted perfume design.

How do you feel after winning this prize? What does it mean for you professionally and personally?

It is, of course, a very good feeling. I have spent over 3 years developing this project and I am very pleased that it is appreciated by the main players in the industry. It gives me energy and motivation to go further and continue what I started years ago.

spot_img
- Sponsorisé -Récupération de DonnèeRécupération de DonnèeRécupération de DonnèeRécupération de Donnèe

Must read

Reportages