Light Field Image Super-Resolution With Transformers

Light field (LF) image super-resolution (SR) aims at reconstructing high-resolution LF images from their low-resolution counterparts. Although CNN-based methods have achieved remarkable performance in SR, these cannot fully model the non-local properties of 4D data. In this paper, we propose a simple but effective Transformer-based method for SR. our method, an angular Transformer is designed to incorporate complementary information among different views, and spatial developed capture both local long-range dependencies within each sub-aperture image. With proposed Transformers, beneficial can be exploited SR boosted. We validate effectiveness Transformers through extensive ablation studies, compare recent state-of-the-art on five public datasets. Our achieves superior with small size low computational cost. Code available https://github.com/ZhengyuLiang24/LFT.