Search papers, labs, and topics across Lattice.
This paper introduces a graph-based learning framework that reconstructs stitching information from 2D sewing patterns, addressing the challenge of disjoint panels without explicit stitch annotations. By predicting panel semantics tied to anatomical regions, the method establishes coarse panel connectivity while utilizing latent edge representations to infer fine-grained seam correspondences through graph message passing. The approach achieves high stitching accuracy and generalizes well across various garment styles, significantly reducing the need for expert input in 3D modeling of sewing patterns.
Achieving high stitching accuracy from 2D patterns without expert annotations could revolutionize the automation of garment design.
Digital sewing patterns typically consist of disjoint 2D panels without explicit stitch annotations, making downstream 3D modeling reliant on labor-intensive expert specification. In this paper, we present a graph-based learning framework that reconstructs two-level stitching information, coarse panel connectivity and fine-grained seam correspondence, from 2D panel geometry alone. At the coarse level, panel connectivity is inferred by predicting panel semantics associated with anatomical body regions, enforcing consistency with body structure and garment design conventions. Based on the reconstructed panel graph, fine-grained seam correspondences between panel pairs are inferred by learning latent edge representations that jointly encode local seam geometry and global garment context through graph message passing. The resulting edge embeddings are subsequently decoded into detailed seam correspondences. Our method supports complex sewing-pattern topologies, including many-to-one correspondences, intra-panel seams, and curved seams. Experiments demonstrate high stitching accuracy and strong generalization across garment styles.
