Domain-Tuned Diffusion Model Upscales Facial Assets to 4K for Cinematic Use
A new arXiv paper presents a domain-tuned diffusion super-resolution method that generates 4K albedo and displacement maps for physically based facial assets, addressing pipelines that typically output only 1K textures.
Read the full story at arXiv cs.CV ↗
Timeline · 1 report
- 2026-11-18 05:00 · arXiv cs.CV
Domain-Tuned Diffusion Super-Resolution for 4K Facial Asset Maps