Bas-relief modelling from enriched detail and geometry with deep normal transfer

Detail-and-geometry richness is essential to bas-relief modelling. However, existing image-based and model-based modelling techniques commonly suffer from detail monotony or geometry loss. In this paper, we introduce a new framework for abundance with visual attention based mask generation preservation, which benefits our two key contributions. For richness, propose novel semantic neural network of normal transfer enrich the texture styles on bas-reliefs. decomposition scheme Domain Transfer Recursive Filter (DTRF). Experimental results demonstrate that approach advantageous producing modellings both fine details preservation.