An Application to Generate Style Guided Compatible Outfit

TL;DR

This paper introduces a novel style encoder network for generating fashion outfits guided by specific styles or themes, advancing the capability of style-aware outfit recommendation systems.

Contribution

The paper presents an end-to-end methodology for style-guided outfit generation using a new style encoder network, with extensive analysis and a demonstration API.

Findings

Effective style-guided outfit generation demonstrated

Analysis of style influence on outfit compatibility

API showcases practical application of the method

Abstract

Fashion recommendation has witnessed a phenomenal growth of research, particularly in the domains of shop-the-look, contextaware outfit creation, personalizing outfit creation etc. Majority of the work in this area focuses on better understanding of the notion of complimentary relationship between lifestyle items. Quite recently, some works have realised that style plays a vital role in fashion, especially in the understanding of compatibility learning and outfit creation. In this paper, we would like to present the end-to-end design of a methodology in which we aim to generate outfits guided by styles or themes using a novel style encoder network. We present an extensive analysis of different aspects of our method through various experiments. We also provide a demonstration api to showcase the ability of our work in generating outfits based on an anchor item and styles.