GarmentImage: Raster Encoding of Garment Sewing Patterns with Diverse Topologies

TL;DR

This paper introduces GarmentImage, a raster-based encoding for garment sewing patterns that improves neural network handling of diverse topologies and enhances generalization in pattern-related tasks.

Contribution

GarmentImage provides a unified raster representation for sewing patterns, enabling neural networks to better handle topology variations and generalize to unseen garment designs.

Findings

Outperforms vector-based methods in pattern exploration and editing

Enables seamless topology transitions in neural representations

Achieves superior results with simple convolutional networks

Abstract

Garment sewing patterns are the design language behind clothing, yet their current vector-based digital representations weren't built with machine learning in mind. Vector-based representation encodes a sewing pattern as a discrete set of panels, each defined as a sequence of lines and curves, stitching information between panels and the placement of each panel around a body. However, this representation causes two major challenges for neural networks: discontinuity in latent space between patterns with different topologies and limited generalization to garments with unseen topologies in the training data. In this work, we introduce GarmentImage, a unified raster-based sewing pattern representation. GarmentImage encodes a garment sewing pattern's geometry, topology and placement into multi-channel regular grids. Machine learning models trained on GarmentImage achieve seamless transitions between patterns with different topologies and show better generalization capabilities compared to models trained on vector-based representation. We demonstrate the effectiveness of GarmentImage across three applications: pattern exploration in latent space, text-based pattern editing, and image-to-pattern prediction. The results show that GarmentImage achieves superior performance on these applications using only simple convolutional networks.