For those seeking an AI perspective on fashion, Colorful Board Inc.’s Sensy app for smart-phones seeks to determine the optimum outfit based on the user’s tastes. “It’s like a fashion coordinator who chooses clothes for me as I’m too busy to spare time for selection,” said Sachi Okuyama, a 36-year-old company employee in Tokyo.
The Sensy app, launched in November, uses an AI system the Tokyo-based information technology venture developed jointly with researchers at Keio and Chiba universities. Users of the service download the free app and sort out whether they like the images of wear sent to their smart-phones once a day. The AI system analyzes replies from the user in accordance with color, shape, price and 47 other criteria to find out that person’s taste, such as “favoring pin-striped red wear of famous brands at sharp discounts.”
Colorful Board has tied up with more than 2,000 fashion companies and online commercial sites both at home and abroad for women in their 20s and 30s. Out of a huge number of dresses introduced on the Internet, the AI system recommends clothes to each user based on accumulated data. When users purchase recommended clothes, sellers pay commissions to Colorful Board.
Okuyama recently bought a gray one-piece suit using Sensy “The more the Sensy app is used, the better matches it can recommend because it learns every day,” said Yuki Watanabe, president of Colorful Board.
It would be interesting if Sachi Okuyama could try out Kokko's ColorSisters we described a month ago. This would inform us whether machine learning can beat an expert system based on the knowledge of experts. I can imagine that the answer will be culture-dependent: Japanese or Germans prefer to blend in and for them an estimate based on big data would be preferable. Americans like to stand out and Italian like to express their individuality, so I can imagine they would prefer the advice of fashion and cosmetics experts.
But then, Sachi Okuyama could be a rebel, so we need plenty of data.
Source: The Japan Times, Artificial intelligence apps guiding users’ clothing choices, culinary tastes.