A Localized AIGC-Live2D Workflow for Dynamic Virtual Human Clothing Generation: Deployment and Validation

Abstract

Live2D creators on the open-source digital human platforms face three core challenges: fragmented workflows, difficult local deployment, and low-efficiency UV mapping. This study proposes a localized AIGC-Live2D workflow with five sequential modules. The full workflow takes 31.2 minutes, improves efficiency by 75% over traditional manual UV mapping, and achieves 4.4/5 satisfaction among 30 professional designers. It solves texture misalignment and deployment barriers, lowers the creation threshold, and provides a replicable scheme for digital media art practitioners.