Designing Woven Patterns with Generative Adversarial Networks: A Comprehensive Survey
Andiko Putro Suryotomo (a*), Bagus Muhammad Akbar (a), Dhimas Arief Dharmawan (a), Nur Indrianti (b), Anindia Az-Zahra (a), Sayang Sani (a), Devina Azzahra (b)

a) Department of Informatics, Universitas Pembangunan Nasional ^Veteran^ Yogyakarta
Jl. Tambak Bayan No.2, Janti, Caturtunggal, Kec. Depok, Kabupaten Sleman, Daerah Istimewa Yogyakarta 55281, Indonesia
*andiko.ps[at]upnyk.ac.id
b) Department of Industrial Engineering, Universitas Pembangunan Nasional ^Veteran^ Yogyakarta
Jl. Tambak Bayan No.2, Janti, Caturtunggal, Kec. Depok, Kabupaten Sleman, Daerah Istimewa Yogyakarta 55281, Indonesia


Abstract

This paper investigates the use of Generative Adversarial Networks (GANs) in designing woven textile patterns, focusing on enhancing creativity and cultural relevance. It examines existing methods like StyleGAN and SinGAN, identifying limitations in generating diverse and emotionally resonant designs. Through experiments with the Ulos motif-a traditional textile from the Batak tribe-this study demonstrates the effectiveness of improved StyleGAN techniques in producing high-quality, varied motifs. The research emphasizes the need to integrate human creativity into GAN-driven processes to avoid homogenization and maintain cultural significance. It also addresses future challenges, including the need for more diverse datasets and ethical considerations in AI-generated design. This work contributes to the understanding of how GANs can be used to balance technical innovation with cultural and emotional depth in textile design.

Keywords: AI-Driven Creativity, Cultural Relevance, Generative Adversarial Networks (GANs), Woven Textile Design

Topic: Engineering

ICARSESS 2024 Conference | Conference Management System