A new EU-founded project is planning on using AI to try and reproduce the key scents and smells of Europe’s history.
The team working on it, named Odeuropa, is on its way to create machine learning and computer vision techniques that can be used to detect information and odors from documents and images for references – such as the smell of tobacco. Then, they will use it to recreate a selection of scents by using heritage science techniques.
Odeuropa aims to re-discover the smells and the stories of our past by uncovering the sensory information from Europe’s history. For this, the team of historians, scientists, and perfumers will get into the digital heritage collections and explore how the scents can be expressed in various places, times, and languages.
Odeuropa wants to develop a ‘computer nose’ that would be able to detect smells and sensory experiences in digital texts over four centuries and seven languages. Then, they would share their findings with museum visitors in a series of public events in the next three years and an online archive will make it accessible to everyone.