High-Fidelity Light Field Reconstruction Method Using View-Selective Angular Feature Extraction

Deep learning (DL) provides an effective approach for light field (LF) reconstruction that aims to synthesize novel views from sparse-sampled views. However, it is challenging address domain asymmetry when adopting spatial-angular interaction LF methods. To solve this problem, a view-selective angular feature extraction block (VS-LFAFE) proposed obtain full-resolution features enumerate whole view directions in macropixel. By applying the VS-LFAFE, method with two subblocks: and fusion upsampling block. Experimental results validate VS-LFAFE effective, can achieve superior performance compared state-of-the-art