DeSIGN: Design Inspiration from Generative Networks

TL;DR

This paper explores the use of various generative network architectures and novel loss functions inspired by Sharma-Mittal divergence to create original fashion designs, with an evaluation protocol combining automatic metrics and human studies.

Contribution

It introduces a new framework for fashion design generation using diverse GAN architectures and a novel loss function based on Sharma-Mittal divergence, along with an evaluation protocol.

Findings

61% of generated images are perceived as human-created

Proposed loss functions outperform existing ones in novelty and likability

Generated designs show high originality and human appreciation

Abstract

Can an algorithm create original and compelling fashion designs to serve as an inspirational assistant? To help answer this question, we design and investigate different image generation models associated with different loss functions to boost creativity in fashion generation. The dimensions of our explorations include: (i) different Generative Adversarial Networks architectures that start from noise vectors to generate fashion items, (ii) novel loss functions that encourage novelty, inspired from Sharma-Mittal divergence, a generalized mutual information measure for the widely used relative entropies such as Kullback-Leibler, and (iii) a generation process following the key elements of fashion design (disentangling shape and texture components). A key challenge of this study is the evaluation of generated designs and the retrieval of best ones, hence we put together an evaluation protocol associating automatic metrics and human experimental studies that we hope will help ease future research. We show that our proposed creativity criterion yield better overall appreciation than the one employed in Creative Adversarial Networks. In the end, about 61% of our images are thought to be created by human designers rather than by a computer while also being considered original per our human subject experiments, and our proposed loss scores the highest compared to existing losses in both novelty and likability.