Segmentation task for fashion and apparel
This addresses a domain-specific challenge for fashion industry professionals, but it appears incremental as it applies existing methods to new data.
The paper tackled the problem of tracking fashion items and combinations by implementing several deep learning architectures on the iMaterialist dataset with 45,000 images and 46 categories, but no concrete results or numbers are provided.
The Fashion Industry is a strong and important industry in the global economy. Globalization has brought fast fashion, quick shifting consumer shopping preferences, more competition, and abundance in fashion shops and retailers, making it more difficult for professionals in the fashion industry to keep track of what fashion items people wear and how they combine them. This paper solves this problem by implementing several Deep Learning Architectures using the iMaterialist dataset consisting of 45,000 images with 46 different clothing and apparel categories.