Duluth Company Launches FashionBrain with Oscars
Fashion software hopes to make things easier for content creators.
DULUTH, Minn.- The 91st Academy Awards showcased stars, their movies, and their fashion.
But for fashion publishers and content creators, sorting through and finding the right images to write stories can be a challenge.
A new software is looking to change that, launched from right here in Duluth–an unlikely team up of the complex world of artificial intelligence and the equally complex world of fashion.
The Oscars are full of different dresses, different hairdos, different jewelry, and so on.
So FashionBrain launched, to help categorize and label images, and even find trends.
FashionBrain has been in the works for almost 3 years, the team right here in Duluth working closely with BrainCreators in the Netherlands.
Together, using the keyword standard Runway Manhattan developed 6 years ago, they worked to improve fashion writing, making it more efficient.
“It’s a perfect symbiosis of tech that is available today, and fashion insides and fashion knowledge that we can contribute and that also only has been available over the last 2–3 years,” Founder and Editorial Director Markus Mueller said.
Here’s how it works.
An image is put into the system, of either one or multiple people, from a service like Reuters, Getty, etc.
Then, the system shows you what it found: the person’s name, their hairstyle, whether it’s over or behind the hears, facial hair, dress type, and so much more.
That way you can find that perfect image for your story.
“To find the right image is the search for the proverbial needle in a haystack,” Mueller said.
“Anybody, whether they are in Duluth, Minnesota or in New York or in Berlin or in London, will be able to retrieve the content that is exactly fitting whatever story they would like to create.”
On Oscar Sunday they only released a small version, to tag about 400 keywords.
But for they said the full service should be able to tag more than 20,000 from an image.
The team hopes to launch more versions throughout the year that will even be able to incorporate natural language processing, to suggest text and research.