Handloom Design Generation Using Generative Networks

2020-10-01

International Conference on Information Photonics

AI Summary

Traditional handwoven textile design suffers from low efficiency and difficulties in preserving ethnic stylistic features. Method: This paper proposes a generative design framework tailored for artisanal textiles. We introduce Neural-Loom—the first domain-specific dataset for handwoven fabrics—and conduct a systematic evaluation of GANs, VAEs, and neural style transfer for pattern generation. Further, we propose a multi-model fusion architecture integrated with a user-perception evaluation mechanism. Contribution/Results: Experiments demonstrate that our generated ethnic-style garment patterns achieve significantly higher user ratings than all baselines (p < 0.01), validating the models’ capacity to effectively capture and creatively reconstruct intricate craftsmanship aesthetics. This work establishes a reproducible data foundation, methodological paradigm, and empirical evidence for intelligent intangible cultural heritage (ICH) textile design.

Technology Category

• Humans and AI: Game Design — Procedural Content Generation & Storytelling
• Natural Language Processing: Generation
• Computer Vision: Generative Adversarial Networks (GANs) for Vision

Application Category

• Economics, Online Markets and Human Computation: Economic ramifications for generative AI infrastructure and applications
• Web Mining and Content Analysis: Web data generation and simulation
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking

Abstract

This paper proposes deep learning techniques of generating designs for clothing, focused on handloom fabric and discusses the associated challenges along with its application. The capability of generative neural network models in understanding artistic designs and synthesizing those is not yet explored well. In this work, multiple methods are employed incorporating the current state of the art generative models and style transfer algorithms to study and observe their performance for the task. The results are then evaluated through user score. This work also provides a new dataset ”Neural-Loom” for the task of the design generation.

Problem

• Generating handloom fabric designs using deep learning
• Exploring generative neural networks for artistic design synthesis
• Evaluating performance of generative models via user scores

Innovation

• Generative neural networks for handloom design synthesis
• Style transfer algorithms enhance artistic design generation
• NeuralLoom dataset introduced for design generation tasks