Researchers have developed MeshQuery, a novel agentic approach for automatic UV unwrapping of 3D models. This method utilizes a vision-language model (VLM) to plan seams based on natural language instructions and domain-specific knowledge. MeshQuery employs a queryable mesh representation and a domain-specific language to express seam plans as executable programs, allowing for iterative refinement based on feedback. The system demonstrates significant improvements over existing baselines in terms of chart count and seam length, with results preferred by professional artists. AI
IMPACT This research could streamline 3D asset creation pipelines by automating UV unwrapping, potentially reducing manual effort for artists.
RANK_REASON The cluster describes a new research paper detailing a novel method for UV parametrization. [lever_c_demoted from research: ic=1 ai=1.0]
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →