Liu, Y.-L.Y.-L.LiuLai, W.-S.W.-S.LaiChen, Y.-S.Y.-S.ChenKao, Y.-L.Y.-L.KaoYang, M.-H.M.-H.YangHuang, J.-B.J.-B.HuangYUNG-YU CHUANG2021-05-052021-05-05202010636919https://www.scopus.com/inward/record.url?eid=2-s2.0-85093944674&partnerID=40&md5=6fc5e4a2b241120636909cf87e645972https://scholars.lib.ntu.edu.tw/handle/123456789/559407Recovering a high dynamic range (HDR) image from a single low dynamic range (LDR) input image is challenging due to missing details in under-/over-exposed regions caused by quantization and saturation of camera sensors. In contrast to existing learning-based methods, our core idea is to incorporate the domain knowledge of the LDR image formation pipeline into our model. We model the HDR-to-LDR image formation pipeline as the (1) dynamic range clipping, (2) non-linear mapping from a camera response function, and (3) quantization. We then propose to learn three specialized CNNs to reverse these steps. By decomposing the problem into specific sub-tasks, we impose effective physical constraints to facilitate the training of individual sub-networks. Finally, we jointly fine-tune the entire model end-to-end to reduce error accumulation. With extensive quantitative and qualitative experiments on diverse image datasets, we demonstrate that the proposed method performs favorably against state-of-the-art single-image HDR reconstruction algorithms. © 2020 IEEE.Cameras; Pattern recognition; Pipelines; Camera response functions; Error accumulation; High dynamic range images; Learning-based methods; Nonlinear mappings; Physical constraints; Qualitative experiments; Reconstruction algorithms; Image reconstructionconference paper10.1109/CVPR42600.2020.001722-s2.0-85093944674