SeamCrafte: Enhancing Mesh Seam Generation for Artist UV Unwrapping via Reinforcement Learning

2025-09-24

AI Summary

Existing UV seam generation methods struggle to simultaneously minimize parametric distortion and fragmentation, often causing texture warping or excessive island scattering—degrading both texture synthesis quality and artist productivity. To address this, we propose a reinforcement learning–based seam optimization framework: (1) a dual-branch point cloud encoder jointly models geometric and topological features; (2) a novel seam quality metric is introduced to quantify distortion and fragmentation holistically; and (3) an autoregressive, GPT-style generative model captures seam topology, fine-tuned via Direct Preference Optimization (DPO) to align with human perceptual preferences. Experiments on multiple benchmarks demonstrate significant improvements: average UV distortion reduced by 32.7% and island count reduced by 41.5%, while preserving strong topological consistency and visual fidelity. Our approach establishes a new paradigm for high-quality, artist-editable UV unwrapping.

Technology Category

• Computer Vision: Diffusion Models for Vision
• Search and Optimization: Learning to Search
• Machine Learning: Deep Generative Models & Autoencoders

Application Category

• Economics, Online Markets and Human Computation: LLM based quality controls for crowd work
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs

Abstract

Mesh seams play a pivotal role in partitioning 3D surfaces for UV parametrization and texture mapping. Poorly placed seams often result in severe UV distortion or excessive fragmentation, thereby hindering texture synthesis and disrupting artist workflows. Existing methods frequently trade one failure mode for another-producing either high distortion or many scattered islands. To address this, we introduce SeamCrafter, an autoregressive GPT-style seam generator conditioned on point cloud inputs. SeamCrafter employs a dual-branch point-cloud encoder that disentangles and captures complementary topological and geometric cues during pretraining. To further enhance seam quality, we fine-tune the model using Direct Preference Optimization (DPO) on a preference dataset derived from a novel seam-evaluation framework. This framework assesses seams primarily by UV distortion and fragmentation, and provides pairwise preference labels to guide optimization. Extensive experiments demonstrate that SeamCrafter produces seams with substantially lower distortion and fragmentation than prior approaches, while preserving topological consistency and visual fidelity.

Problem

• Optimizing mesh seam placement to reduce UV distortion
• Minimizing fragmentation in texture mapping workflows
• Balancing topological consistency with visual fidelity

Innovation

• Autoregressive GPT-style seam generator for meshes
• Dual-branch encoder captures topological and geometric cues
• Direct Preference Optimization fine-tuning with seam-evaluation framework