GeoStyle: Discovering Fashion Trends and Events

TL;DR

This paper presents a framework that analyzes street imagery to discover, forecast, and identify fashion trends and events, significantly improving trend prediction accuracy and revealing impactful social events worldwide.

Contribution

It introduces an automatic method for analyzing large-scale street images to uncover and predict fashion trends and detect influential social events, advancing prior trend analysis techniques.

Findings

Trend forecasts are over 20% more accurate than previous methods.

The framework identifies hundreds of socially meaningful fashion-related events.

It effectively analyzes spatial and temporal patterns in fashion data.

Abstract

Understanding fashion styles and trends is of great potential interest to retailers and consumers alike. The photos people upload to social media are a historical and public data source of how people dress across the world and at different times. While we now have tools to automatically recognize the clothing and style attributes of what people are wearing in these photographs, we lack the ability to analyze spatial and temporal trends in these attributes or make predictions about the future. In this paper, we address this need by providing an automatic framework that analyzes large corpora of street imagery to (a) discover and forecast long-term trends of various fashion attributes as well as automatically discovered styles, and (b) identify spatio-temporally localized events that affect what people wear. We show that our framework makes long term trend forecasts that are >20% more accurate than the prior art, and identifies hundreds of socially meaningful events that impact fashion across the globe.