C^2RV: Cross-Regional and Cross-View Learning for Sparse-View CBCT Reconstruction | Read Paper on Bytez