IBM and Marchesa Will Debut 'Cognitive Dress' at Met Gala
IBM and fashion designer Marchesa will debut a "cognitive dress" at the Met Gala Monday night, extending IBM's cognitive computing branding campaign.
The data-driven dress -- which will change colors based on an analysis of fan sentiment in real time -- will be revealed at 7 p.m. ET tonight on the red carpet of the Met Gala event in New York, called "Manus x Machina."
It was designed using Watson artificial intelligence technology, cognitive tools from IBM Research, tech solutions from Watson development partner Inno360, and creative vision from the Marchesa design team.
"This all started when we learned about the Met Gala theme -- 'Man + Machine'," said Ann Rubin, VP-branded content and global creative at IBM. "That is how we talk about Watson -- using cognitive technology to help humans do things better."
Watson technology was used in three ways to create the dress.
First, the Marchesa design team selected five emotions they wanted the dress to convey -- joy, passion, excitement, encouragement and curiosity. A team at IBM Research then fed this data into a cognitive color design tool to come up with colors that were aligned with these emotions.
"Watson used various capabilities to understand how fans react to differ color palettes," Ms. Rubin said.
Once the colors were selected, IBM turned to software company Inno360 to source a fabric for the dress, which uses LED technology to change colors. Important considerations for the fabric were weight, luminosity and flexibility.
Inno360 used different Watson technologies to sort through 40,000 sources of fabric, narrowing down the list to 35 that it gave to the Marchesa team to make the final selection.
The third piece is the data-driven aspect of the dress, which uses Watson Tone Analyzer to tap into social sentiment from Twitter viewers. The dress, which is embedded with LED lights, will change colors in real time based on social conversations.
"This is a great opportunity to use a cultural moment to show people what Watson can do," Ms. Rubin said.
"Now, you can see how fashion designers could use some of these APIs in design, using social sentiment to create color palettes," Ms. Rubin said.