Files in this item



application/pdfLIU-THESIS-2020.pdf (16MB)Restricted Access
(no description provided)PDF


Title:Coherent and controllable outfit generation
Author(s):Liu, Chen
Advisor(s):Forsyth, David
Department / Program:Electrical & Computer Eng
Discipline:Electrical & Computer Engr
Degree Granting Institution:University of Illinois at Urbana-Champaign
Subject(s):Computer Vision
Fashion Compatibility
Image Embedding
Outfit Generation
Abstract:People often select outfits with a theme in mind. One might dress for a tropical getaway or to look good at a cocktail party. Existing outfit generation methods use item-wise or outfit-item compatibility tests, and lack an effective method to enforce a global constraint like style or occasion. We describe the first outfit generation method that can produce an outfit consisting of compatible items that cohere to follow a theme specified by the user. Our method generates outfits whose items match a theme described by a query sentence. Our method uses text and image embeddings to represent fashion items. We learn a multimodal embedding where the image representation for an item is close to its text representation, and use this embedding to measure item-query coherence. We then use a discriminator to compute compatibility between fashion items. This strategy yields a compatibility prediction method that meets or exceeds the state of the art. Our generation method combines item-item compatibility and item-query coherence to construct an outfit whose items are (a) close to the query and (b) compatible with one another. Quantitative evaluation shows that the items in our outfits are tightly clustered compared to standard outfits. Furthermore, outfits produced by similar queries are close to one another, and outfits produced by very different queries are far apart. Qualitative evaluation shows that our method responds well to queries. A user study suggests that people understand the match between the queries and the outfits produced by our method.
Issue Date:2020-07-22
Rights Information:Copyright 2020 Chen Liu
Date Available in IDEALS:2020-10-07
Date Deposited:2020-08

This item appears in the following Collection(s)

Item Statistics