FashionBot
Introduction
The main purpose of this project is to recommend outfits that fit a user’s style, needs, and wants. We aim to style and combine garments that users currently own for their daily needs, and create a more robust clothing recommendation system that stores may implement as well. Beyond its use as a daily tool to assist in choosing outfits, it can also recommend new clothing articles that may fit with users’ existing closets.
This project is unique from existing product recommendation as it attempts to discern the relationships between articles of clothing and recommend combinations of them. Solutions to this specific problem that exist at the moment appear to be underdeveloped. We believe that the success and feasibility of this project also depends on its ability to continuously learn and be fed new data, both from specifically curated datasets and a user’s own outfits, which we have yet to see. This way the model can update itself on the latest trends and create a historical archive as well.
