Research talk: Capturing the visual evolution of fashion in space and time

Опубликовано 25 января 2022, 1:40

Speaker: Kristen Grauman, Professor, University of Texas at Austin

The fashion domain is a magnet for computer vision. New vision problems are emerging in step with the fashion industry's rapid evolution towards an online, social, and personalized business. Style models, trend forecasting, and recommendation all require visual understanding with rich detail and subtlety. Not only can this visual understanding benefit individual users, but when analyzed across large-scale multi-modal data, it also can reveal how cultural factors and world events dynamically influence what people around the world wear. I will present our work investigating fashion forecasting and influence from photos. We introduce models to quantify which cultural factors (as captured by millions of news articles) most affect the clothes people chose to wear across a century of vintage clothing photos, as well as models that discover from web photos the way styles propagate from one city to another over time. This work is a first step towards data-driven, quantitative understanding of how our clothes reflect our ever-changing culture and our interconnected world.

Learn more about the 2021 Microsoft Research Summit: Aka.ms/researchsummit