Canadian Manufacturing

XNFY Lab announce the launch of an AI generated fashion

A powerful and adaptable technology to revolutionize digital brand content and start a new era in the upcycling of garments

April 27, 2021  by CM Staff

Our IP is a patent-pending technology that generates new synthetic original images that can pass for real. It does not copy images or movements, it learns, what a dress is, how it fits and what details can be done to create original content.

VIANA DO CASTELO, Portugal — The fashion industry is based on cycles, trends, and creative/self-expression and like few others is all about visuals, looks and style.

Understanding the current state of fashion and the potential of AI generated content to solve some of its main problems, the XNFY Lab research team developed an IP technology integrating Microsoft Azure Machine Learning, an AI solution, to provide designers, brands and marketplaces a tool that learns from multiple sources and generates high-definition clothing images or professional digital style content. The technology capacity can also be used to “re-think” and generate new design ideas based on unsold inventory and deadstock fabrics for example.

With this technology, brands can show clothes images to their customers via e-commerce or social networks first, even before they exist. That will be a very accurate way to predict the demand of that item and avoid the current status-quo and industry processes, which generates a lot of unsold inventory.

“The technology we are building can be applied in almost every industry, but for fashion we solve real problems and have a positive impact,” said Orlando Ribas, CEO XNFY Lab in a prepared statement.

Advertisment

The vision behind this project is also aligned with the recent Microsoft in Culture, a Microsoft initiative that merges creative industries with the technology to innovate.

Overall, Microsoft Azure Machine Learning enhances the business potential for fashion by empowering researchers to collaborate, improve, train and deploy AI models and accelerate our time-to-market. In addition, it allows customers to validate and test datasets quicker, bringing results in hours versus instead of days.

“We reached the first milestone of our roadmap with impressive results. To celebrate this milestone we are preparing a small, limited edition crowdfunding, where we will put AI design product in the hands of customers,” added Ribas.