Zanyi Wang

From RGB Generation to Dense Field Readout: Pixel-Space Dense Prediction with Text-to-Image Models

Zanyi Wang, Xin Lin, Haodong Li, Dengyang Jiang, Yijiang Li

Large-scale text-to-image models are attractive backbones for dense prediction because RGB generation pretraining learns rich semantic, structural, and geometric priors. Existing generative and editing approaches reuse these priors by casting dense prediction as target generation…Read more